ELSA hints GT206 and GT212

[RussSchultz]

How sure are you about that?

Those seem awfully busy and large to be pads and drivers for 4 pins for each square.

[fellix]

LOL, you obviously haven't seen what a truly large pad array looks like. :lol:

[Jawed]

How sure are you about that?
Compare with the ATI die, where exceedingly similar regions of the die (which are also of corresponding size) are explicitly labelled as memory interface:

http://www.techreport.com/articles.x/14990/2

Oh and look at the layout of signal pads and the plethora of ground and power lines in figure 26:

Under Bump Routing Layer Method and Apparatus (http://forum.beyond3d.com/publicationDetails/biblio?adjacent=true&locale=en_EP&FT=D&date=20090205&CC=US&NR=2009032941A1&KC=A1)

have to admit I haven't read it through.

Jawed

[CarstenS]

I think fellix's ideas are on the right track, though the "stacked" stuff is quite a puzzler.

Carsten if you compare with the annotated GT200 here (even if there are some who doubt its accuracy):

http://www.techreport.com/articles.x/14934/2

you should see that TMUs and ROPs take up acres of space.

Jawed

That shot doesn't distinguish at all between SIMD-Control- and TU-logic - it's all "Texture", whereas the GT21x-shot at least shows some additional logic besides the actual ALUs in the SIMD-parts of the die - but just not enough to make me believe, that TUs are still incorporated.

I'm not saying, fellix is wrong and I am here, but I've yet to see convincing evidence for either position.

And take into account, that for DX10.1 Nvidia would have to overhaul their TMUs either way - and maybe they tried to get away with less space, maybe combining some of the stuff for accessing memory, which is replicated in both TMUs and ROPs.

I could imagine, you can get away with less space when routing a dual-lane (1 for ROP-use, 1 for TU-use) to memory compared to having to to the individual routing from two far away parts of the die (I guess that's the principle of highways or autobahns also).

[fellix]

There are four similar structured rectangle blocks, situated between the pairs of TPCs distinguishable in the die shot -- those could be texturing hardware, being just the samplers, mapping units or even both (too small for eight TMU quads, anyway... duh!). :???:
Just to make my statement more figurative (the red outline):

http://img8.imageshack.us/img8/3817/gt215.png

[Jawed]

That shot doesn't distinguish at all between SIMD-Control- and TU-logic - it's all "Texture", whereas the GT21x-shot at least shows some additional logic besides the actual ALUs in the SIMD-parts of the die - but just not enough to make me believe, that TUs are still incorporated.
Try this, too:

http://pc.watch.impress.co.jp/docs/2008/0617/kaigai_16l.gif

Even though it, too, doesn't make the distinctions you require.

I agree there should be some control stuff per cluster and I've no idea of the extent of the SIMD-specific stuff (i.e. 3x MAD-8, MI-2 and DP-1).

And take into account, that for DX10.1 Nvidia would have to overhaul their TMUs either way - and maybe they tried to get away with less space, maybe combining some of the stuff for accessing memory, which is replicated in both TMUs and ROPs.
Yes, they definitely have to do extra things (e.g. gather). We still don't even know how many TMUs there are. For all we know there's only 16 of them :razz:

I could imagine, you can get away with less space when routing a dual-lane (1 for ROP-use, 1 for TU-use) to memory compared to having to to the individual routing from two far away parts of the die (I guess that's the principle of highways or autobahns also).
Yes, to a degree "repeater islands" across the die imply that routing will agglomerate. The routes themselves don't take space since they are in metal layers under the logic layer.

Jawed

[Novum]

What you've marked doesn't add up to an entire cluster. Could be general control or it could be TMU. Dunno.
What's missing? NVIDIA marks it the same.

Just to make my statement more figurative (the red outline):

http://img8.imageshack.us/img8/3817/gt215.png
Hrm that could explain it. But then that NVIDIA picture is wrong:

http://images.anandtech.com/reviews/video/ATI/4800/gt200.jpg

There should be more "random" logic that is not texture that belongs to the ALUs.

[RussSchultz]

Just to make my statement more figurative (the red outline):

http://img8.imageshack.us/img8/3817/gt215.png
The red squares don't look to be instances. The contents look similar, but they don't look like instances.

And the blue squares seem to be too big. (the areas closer to the center line do not match across instances)

[fellix]

I think we've already concluded here, that irregularities between similar block instances are due to employing a full automatic design & tuning for the selected logic circuits.

[RussSchultz]

What you're pointing to doesn't look like a sea of gates, either (i.e. the product of automatic place and route).

I mean, I guess it just doesn't make sense to me to 'halfway instance' something in a way that looks close to the same, but not quite.

Usually instancing is either plopping hard macros down, or just letting the auto place and route do its thing and ending up with a sea of gates.

Of course, I only tangentially work in the back end of chip design, so I just might not be familiar with the technique.

[Jawed]

http://www.digitimes.com/news/a20090707PB203.html
Nvidia expects to launch its new 40nm GeForce GT220 (GT216) and GeForce G210 (GT218) GPUs at the end of September 2009, according to a Chinese-language Commercial Times report citing sources from graphics card makers.

"Low yields on Taiwan Semiconductor Manufacturing Company's (TSMC's) 40nm process caused Nvidia to postpone the launch of its 40nm GPU, but recently TSMC's yield rate has started to improve prompting Nvidia to launch the GPUs in time for the year-end holiday sales, added the paper."

[DSC]

http://www.nvidia.com/object/product_geforce_gt_220_us.html

http://www.nvidia.com/object/product_geforce_g210_us.html

[willardjuice]

http://www.nvidia.com/object/product_geforce_gt_220_us.html

http://www.nvidia.com/object/product_geforce_g210_us.html

Both list DirectX 10.1 support...

[Jawed]

Still says, seemingly erroneously, "16 cores" for the smallest of these.

And the missing GT215 appears to imply that was just a paper launch in mobile form.

Jawed

[Kaotik]

Both list DirectX 10.1 support...

Is that some kinda surprise? The mobile parts, based on same chips, already told us that.

[no-X]

nVidia isn't in good situation. Should they promote DX10.1 (because it's the only new feature, which this 40nm half-gen introduced) and admit, that their entire desktop product range is somewhat obsolete, or protect sales of desktop products, not to promote DX10.1 and give up the only new marketing/product-selling feature?

[kemosabe]

Seems they only needed to promote DX10.1 behind closed doors since according to NVDA management they've already secured a large number of OEM design wins for the notebook market. That should allow them to continue to merrily neglect the feature in the retail desktop space.

[willardjuice]

Is that some kinda surprise? The mobile parts, based on same chips, already told us that.

No but it's profound. Nvidia's first desktop DirectX 10.1 products are worth noting.

[Jawed]

http://www.shopping.hp.com/webapp/shopping/computer_series.do?storeName=computer_store&category=desktops&series_name=p6110t_series&a1=Category&v1=Everyday%20computing

The G210 is $10 cheaper than the HD4350 as an option, or $50 more than Intel integrated graphics.

Jawed

[AnarchX]

http://www.shopping.hp.com/webapp/shopping/computer_series.do?storeName=computer_store&category=desktops&series_name=p6110t_series&a1=Category&v1=Everyday%20computing

The G210 is $10 cheaper than the HD4350 as an option, or $50 more than Intel integrated graphics.

Jawed
http://www.shopping.hp.com/webapp/shopping/computer_can_series.do?storeName=computer_store&category=desktops&a1=From+price&v1=Over+%24600&series_name=e9150t_series

GT 220: $160 over HD4350, $40 below GTS 250 :lol2:

[Tchock]

Edit: whoops brainfart.



HP quotes a build date of July 14 for the G210 systems.


Anyone want to buy one and see what it really is? :smile:

[Vincent]

Edit: whoops brainfart.



HP quotes a build date of July 14 for the G210 systems.


Anyone want to buy one and see what it really is? :smile:



Crysis---- 0-4 Fps :razz:

[neliz]

that G210, isn't that the rebranded G110

16 cores

http://www.nvidia.com/object/product_geforce_g110m_us.html
http://www.nvidia.com/object/product_geforce_g210_us.html

[Kaotik]

that G210, isn't that the rebranded G110

16 cores

http://www.nvidia.com/object/product_geforce_g110m_us.html
http://www.nvidia.com/object/product_geforce_g210_us.html

16 cores in both, sure, but 210 has 10.1 support (and is made on 40nm process)

[rjc]

Kindof messed up report at expreview: Nvidias GT230 to debut in December (http://en.expreview.com/2009/07/23/nvidias-gt230-to-debut-in-december.html)

Designed for mainstream market, GT230 is said to feature GDDR5/GDDR3 memory, and the number of stream processors may reach 96, or just 64 or 80 ones.


It should probably be the first GT215 based part, which on desktop should be labelled something like GTS2X0 where X=3,4.

[Tchock]

GT230 might be male, female, or just bi. :lol:

Makes sense that they might want to castrate a bit for yields, since 140/240's performance at bulk OEM prices would leave demand to solutions that use less power and have a smaller TDP instead.

[CarstenS]

Do we still want DX10(.1) chips in December?

[Kaotik]

Do we still want DX10(.1) chips in December?

If they have nothing else to offer, it's better than no new chips at all for many

[rjc]

GT230 might be male, female, or just bi. :lol:

Makes sense that they might want to castrate a bit for yields, since 140/240's performance at bulk OEM prices would leave demand to solutions that use less power and have a smaller TDP instead.

Yes not sure what they are planning for the clocks on the desktop but if the 96sp version is aggressively clocked may go close to needing a power adapter. Lower specced models obviously not.

Nvidia are much keener on this in the mobile space where it will go up against the RV740. Like the RV740 only think will go into fairly upmarket models....too much heat and power for the multitude of CULV based stuff coming for the rest of the year.

At the conference call the other day AMD said there would be no DX11 notebook models from them till next year. For laptops with discrete gpu this christmas will have a choice of something like: RV710 vs GT218 (60% of models on sale), RV730 vs GT216 (30% of models) and RV740 vs GT215 (10% of models) with IGP obviously alot more.

[Tchock]

GT210M is a really nice improvement across G110M that I think will catch on nicely for the holidays, but GT240M seems to be a really marginal increase even with the enhanced unit count wrt GT130M.

ALUs: 48 // 32
Core: 550//600
Shader: 1210//1400
MEM: 800//1016

Fillrates will benefit, but shader throughput is 1.3x, and total bandwidth is lower :o

I'd reckon that the Mob 4650 would actually win in this segment, so it's up to prices and driver support. Mob 4670 is outclassed by the GT215 (250M) which is 28W+RAM.

RV740 would have a problem if it still can't do a seamless memory power state transition.



After the GF8600M incident, OEMs have been much less enthusiastic on using midrange mobile GPUs I suppose. The focus is mostly at the lowend discrete market instead, since both run The Sims 3 adequately w/o issues. :lol:

[trinibwoy]

GTS 240 just popped up under Nvidia's OEM only section. Yes, it's just a renamed 9800 GT. :evil:

http://www.nvidia.com/object/product_geforce_gts_240_us.html

[mczak]

GTS 240 just popped up under Nvidia's OEM only section. Yes, it's just a renamed 9800 GT. :evil:

http://www.nvidia.com/object/product_geforce_gts_240_us.html
Not quite. It's old chip but it's not only rename. It's got beefed up clocks - +12% core (675 instead of 600 Mhz), +8% shader (1620 vs 1500 Mhz), +22% memory (1100 vs 900 Mhz). In fact that's almost exactly (except shader clock) the clocks of 9800GTX. Should be faster than 9800GT and I guess might even beat an old 8800GTS. Still, that chip is getting old :-).
Oh and btw is this a new board layout?

[rjc]

Kindof messed up report at expreview: Nvidias GT230 to debut in December (http://en.expreview.com/2009/07/23/nvidias-gt230-to-debut-in-december.html)

Apologies to the Expreview people! Looks like they were right with the GT230 name.

Is sort of a bit messy, nvidia kindof needs one more prefix ie:
G210 GT218
GT220 GT216
GT230 GT215
GTS240 G92
GTS250 G92

The G92 has taken 2 spots, and forced the GT215 and GT216 to share a prefix. Also makes it a bit messy for speed/unit count/memory type variations as will now need to use the last digit extensively or limit the variations available. ie DDR3 and GDDR5 versions of GT215 which would likely have quite a performance delta can only really be separated by the last digit.

Figure they were trying to get a GDDR5 version up within sight of the 9800gt performance but couldnt get there, so had to bring in the hardened veteran back off of the bench....:wink:

(...or possibly they had huge stocks of 9800gt and needed few months longer to clear)

[mczak]

Figure they were trying to get a GDDR5 version up within sight of the 9800gt performance but couldnt get there, so had to bring in the hardened veteran back off of the bench....:wink:

(...or possibly they had huge stocks of 9800gt and needed few months longer to clear)

Well, I think if you accept that nvidia wants to ship GTS250 for a while longer, it makes perfect sense that there's also a GTS240. Presumably there are still parts with defects which otherwise nvidia couldn't sell (or had to sell as the "older" 9800GT). The higher clocks also make perfect sense (I bet any g92b can reach that easily, plus there's no danger of hurting sales of original 9800GTX any more), though higher memory clock might potentially drive up price a bit (but maybe not, I don't think those 1.1Ghz are actually higher binned quality, last time I checked everything over 1.0Ghz was still overvolted, plus now there's 1.3Ghz gddr3 parts available so it's no longer the premium part).
Also rumours for GT215 still vary wildly, so it's totally unclear how close to G92 in performance it could be even in the top configuration. I guess with 96 shaders it could get very close, but anything less and I'm sceptical...

[trinibwoy]

The thing is that G92b has some pretty wicked clocks right now - 740 core, 1840 shader. A 96 shader part with only 32 TMUs would probably have a pretty tough time catching up even with GT2xxs enhancements. They were probably looking at GT214 as a true G92 replacement.

[Vincent]

The thing is that G92b has some pretty wicked clocks right now - 740 core, 1840 shader. A 96 shader part with only 32 TMUs would probably have a pretty tough time catching up even with GT2xxs enhancements. They were probably looking at GT214 as a true G92 replacement.


Last time from what I heard that all GT21X series are for low-end.

To compete RV870XT , 2 X G92 on the Level of performance is required.

[trinibwoy]

Yeah I'm talking about the $100-$150 market that G92b now occupies.

[rjc]

Some GT216 shots and benchmarks have turned up at IT168:
http://publish.it168.com/2009/0807/20090807051201.shtml

Core shot directly linked here (http://img.publish.it168.com/2009/0724/images/1601715.jpg)

...competitive with RV730, quite a bit above the 9500.

Two configs:
GT220: 48sp 625/1360/1600mhz DDR3 either 512mb or 1024mb
or
GT215: DDR2 memory (clocks and shaders not listed)

In tests they compare to a RV730, and also a G210 at 589/1402/1000mhz

From the nvidia slide, GT220 comes in 50W and 75W versions(guessing this is for the different memory levels).

(Please note in the above the GT215 is not the GT215 we have come to know and love, instead is a nvidia retail name for the lower clocked GT216) :wink:

[Dave Baumann]

That is a bandwidth choked HD 4670 configuration they have chosen there.

[trinibwoy]

(Please note in the above the GT215 is not the GT215 we have come to know and love, instead is a nvidia retail name for the lower clocked GT216) :wink:

Heh, that's really annoying for those of us who care about this stuff :)

[entity279]

@Dave
Why do you say that? Isn't 1000 MHz (2000 effective) the standard for GDDR3 HD4670s ?

[Dave Baumann]

@Dave
Why do you say that? Isn't 1000 MHz (2000 effective) the standard for GDDR3 HD4670s ?
Looking at the other cardx in that test that appears to be the effective rate, in other words the HD 4670 is 500Mhz on the memory. Look at The Vantage scores for HD 4670 on its release memory clocks and you'll see the scores are significantly higher than tested here.

[no-X]

I tried to complete the list of "GeForce 100/200" graphics cards:

G92b -> GeForce GTS240 (112SPs, 256bit GDDR3, 675/1620/2200), 55nm, DX10 (GF8800GT/GF9800GT/GTS240 OEM)
G92b -> GeGorce GTS150 (128SPs, 256bit GDDR3, 740/1835/2000), 55nm, DX10 (GF8800-512/GF9800GTX/GF9800GTX+/GTS250)
G92b -> GeForce GTS250 (128SPs, 256bit GDDR3, 740/1835/2200), 55nm, DX10 (GF8800-512/GF9800GTX/GF9800GTX+/GTS250)
G94 -> GeForce GT130 (48SPs, 192bit DDR2, 500/1250/1000) 55-65nm, DX10 (crippled 94-based 9600GSO-512)
G98 -> GeForce G100 (8SPs, 64bit DDR2, 565/1400/1000) 55-65nm, DX10 (crippled GF9400GS)
G96 -> GeForce GT120 (32SPs, 128bit DDR2, 500/1400/1000) 55-65nm, DX10 (GF9400GT)
GT200 -> GeForce GTX260 (192SPs, 448bit GDDR3, 575/1240/2000), 65nm, DX10
GT200 -> GeForce GTX260-216 (216SPs, 448bit GDDR3, 575/1240/2000), 65nm, DX10
GT200 -> GeForce GTX280 (240SPs, 512bit GDDR3, 600/1300/2200), 65nm, DX10
GT200b -> GeForce GTX275 (240SPs, 448bit GDDR3, 635/1405/2000), 55nm, DX10
GT200b -> GeForce GTX285 (240SPs, 512bit GDDR3, 650/1475/2280), 55nm, DX10
GT206 -> ??? canceled?, 40nm, DX10.1
GT212 -> ??? (240SPs, 256bit GDDR5, ???), 40nm, DX10.1 - canceled?
GT214 -> on the way, details unavailable, sources: Fudzilla (http://www.fudzilla.com/index.php?option=com_content&task=view&id=11788&Itemid=1), Electronista (http://www.electronista.com/articles/08/12/29/nvidia.gt214.gt216.chips/)
GT215 -> on the way, GeForce GT230 (64-96SPs, 128bit (G)DDR3/5, ???), 40nm, DX10.1 (shrinked G92?)
GT216 -> GeForce GT220 (48SPs, 128bit DDR3, 615/1335/1580), 40nm, DX10.1 shrinked G94?)
GT218 -> GeForce G210 (16SPs, 64bit DDR2, 590/1400/1000), 40nm DX10.1 (shrinked G96/9400GT)


Many articles states, that GT130 is renamed 8800GS/9600GSO (G92). That's not true, GT130 doesn't have 96SPs, only 48. It's based on G94 - it's G94 based 9600GSO-512 with narrower memory bus.

[Lukfi]

Isn't GeForce G100 based on the G98 core (8 SPs, 8 TUs, 64bit memory)?

[no-X]

Thanks, fixed.

[mczak]

Looking at the other cardx in that test that appears to be the effective rate, in other words the HD 4670 is 500Mhz on the memory. Look at The Vantage scores for HD 4670 on its release memory clocks and you'll see the scores are significantly higher than tested here.
Seems you're right. Seems like that GT220 is in fact, not quite competitive with a real HD4670 (be it the 1Ghz GDDR3 or 873Mhz DDR3 version). I think though that's a pretty sleazy HD4670, and I still don't quite understand why AMD allows its partners to do such stunts (it might have chip clock like a HD4670 but nevertheless will perform almost exactly like a HD4650 - in fact there exist HD4650 with gddr3 memory which will be a lot faster than this so-called HD4670 with gddr2 or whatever memory).

quite a bit above the 9500
where did you see that? I tried to look up some scores for the 9500GT and it seems pretty much dead even in performance (though, comparing to the 9500GT GDDR3 version not the much slower ddr2 version). That's a bit disappointing imho, though if it actually sells with ddr3 ram it is indeed quite an improvement (reviews of 9500GT may use gddr3 ram but what you'll find on shelves is typically the ddr2 version...).

[rjc]

GT214 -> on the way, details unavailable, sources: Fudzilla (http://www.fudzilla.com/index.php?option=com_content&task=view&id=11788&Itemid=1), Electronista (http://www.electronista.com/articles/08/12/29/nvidia.gt214.gt216.chips/)
GT215 -> on the way, GeForce GT230 (64-96SPs, 128bit (G)DDR3/5, ???), 40nm, DX10.1 (shrinked G92?)
....
Many articles states, that GT130 is renamed 8800GS/9600GSO (G92). That's not true, GT130 doesn't have 96SPs, only 48. It's based on G94 - it's G94 based 9600GSO-512 with narrower memory bus.
Am pretty sure the GT214 and GT215 are very closely related in functionality ie variations....apparently are trying to get going a UMC as well as TSMC also GT214 might only have had a 192bit interface GDDR3/DDR3 only, not 100%. Really would not expect the GT214 at all now, GT215 will be available in a few months and will sell to say roughly middle of next year when the GT30X parts should arrive.

The 9600GSO was originally a cut down G92(192bit) that started life as the 8800GS and was renamed to fit into the 9 series....nvidia i guess with better yields on the G92 abandoned this part and decided cut down and rename the 9600GT/G94 with the same name to replace the original renamed part. It was quite a bit less performance than the original 9600GSO.

Sorry, was reading too fast, is the GT130 not about the 9600GSO - the GT130 is G94 only i think.....everytime i see the 9600GSO i cant seem to stop myself trying to fit "rename" into a sentence as many times as possible :wink:

[rjc]

where did you see that? I tried to look up some scores for the 9500GT and it seems pretty much dead even in performance (though, comparing to the 9500GT GDDR3 version not the much slower ddr2 version). That's a bit disappointing imho, though if it actually sells with ddr3 ram it is indeed quite an improvement (reviews of 9500GT may use gddr3 ram but what you'll find on shelves is typically the ddr2 version...).

Oh sorry, took the article at face value, they said in chinese:

GeforceGT220并非如此简单，在性能大大超越Geforce9500GT

Translated as usual roughly:
"Geforce GT220 isn't so simple, in performance greatly exceed Geforce 9500GT"

With all these cards the margins are so tight the incentive to "cheat" by substituting memory is overwhelming. The only hope currently have is the price of DDR3 continues its downward march and gets close soon to the currently rising DDR2 so that there is no $ incentive to use the older memory.

[Dave Baumann]

Seems you're right. Seems like that GT220 is in fact, not quite competitive with a real HD4670 (be it the 1Ghz GDDR3 or 873Mhz DDR3 version). I think though that's a pretty sleazy HD4670, and I still don't quite understand why AMD allows its partners to do such stunts (it might have chip clock like a HD4670 but nevertheless will perform almost exactly like a HD4650 - in fact there exist HD4650 with gddr3 memory which will be a lot faster than this so-called HD4670 with gddr2 or whatever memory).
I don't know what all the partners are doing with HD 4600, but I'd like to know what board was in that test. Looking at Newegg the minimum speeds for HD 4670 available there at the moment appear to 800MHz (there was one badly labelled Sapphire board, but referencing against their site indicates is actuall >1600Mbps data rate), while 500Mhz is on the HD 4650.

[mczak]

I don't know what all the partners are doing with HD 4600, but I'd like to know what board was in that test. Looking at Newegg the minimum speeds for HD 4670 available there at the moment appear to 800MHz (there was one badly labelled Sapphire board, but referencing against their site indicates is actuall >1600Mbps data rate), while 500Mhz is on the HD 4650.
I wasn't able to find a HD4670 with 500Mhz ram neither - found some references with HD4670 and gddr2 ram but these all seemed to be misprints. Must be some version only found in China.

[Jawed]

http://img199.imageshack.us/img199/9731/p8266066.th.jpg (http://img199.imageshack.us/i/p8266066.jpg/)
Is that a "week 32 of 2008" manufacturing date? 13 months ago?

Jawed

[no-X]

Maybe they used the GT230 / 192bit board to clean old stock of GPUs which have broken memory controller and/or block of ROPs.

[Jawed]

http://www.fudzilla.com/content/view/15881/34/

[Blazkowicz]

gigabyte will launch a GT220, with dual slot cooler, 1GB of DDR3 and nice clocks.
It looks like an excellent solution for some needs, couple it with an athlon X4 605e and you get very high perf/watt for very cheap.

the only problem is competition from above. here's a 9800GT green selling for really cheap
http://www.grosbill.com/4-gainward_9800gt_512_green_0568_-94861-informatique-nvidia

[Tchock]

GT240 // GTS 260 M (same specs almost exact clocks).
http://image3.it168.com//2009/10/9/4ddd891d-bca7-4ed6-8adc-c01a0459fdd4.jpg
http://image3.it168.com//2009/10/9/ad383fa4-8409-4bee-b141-cd1bfabe8073.jpg
http://image3.it168.com//2009/10/9/a2fa0a36-6dda-4de2-9041-46497fdb9fa9.jpg

[trinibwoy]

Are those numbers legit? Seems strange that it would beat a 250 in Vantage with lower theoretical numbers in everything, especially after being beaten by the same card rather soundly in 3dmark06.

[Arty]

Are those numbers legit? Seems strange that it would beat a 250 in Vantage with lower theoretical numbers in everything, especially after being beaten by the same card rather soundly in 3dmark06.
That is the OC number you are looking at. It is still slower than the 250, even the GDDR5 version.

Is this supposed to go up against the 5700? Doesnt bode well ..

[no-X]

It seems it will be slower than HD4770 (which is 2-3% slower than HD4850 in Vantage) and slightly faster than HD4830 (15-18% slower than HD4850).

[trinibwoy]

That is the OC number you are looking at. It is still slower than the 250, even the GDDR5 version.

I know, it was the OC I was comparing. Even overclocked it has lower theoretical specs than the 250 and the 250 beats the OC soundly in 3dmark06.

[AnarchX]

Probably only 8 ROPs and not 8 per 64-bit MC like Fermi.

[elsence]

Geforce G210 review:

http://www.pcgameshardware.com/aid,694223/Geforce-G210-Nvidias-first-DirectX-101-card-reviewed/Reviews/


Geforce G220 review:

http://publish.it168.com/2009/0901/20090901012701.shtml


Geforce G240 preview:

http://vga.it168.com/a2009/1009/754/000000754803.shtml

General info/photos

http://www.mymypc.com/images/vga/mymypc-nvidia-gt210-gt220.gif

http://www.tcmagazine.com/comments.php?shownews=30077&catid=2

EDIT*

Tomshardware, Pcper and hardwarezone review: (links from: http://www.elitebastards.com)

http://www.pcper.com/article.php?aid=794

http://www.hardwarezone.com/articles/view.php?cid=3&id=3035

http://www.tomshardware.co.uk/geforce-gt-220,review-31703.html

[AnarchX]

D3D10.1 seems buggy:http://www.tomshardware.com/reviews/geforce-gt-220,2445-13.html

[Jawed]

Yay, we finally know the TMU and ROP counts: 16,8 for GT216 and 8,4 for GT218.

http://www.semiaccurate.com/2009/10/11/nvida-launches-uncompetitive-g210-and-gt220/

Jawed

[mczak]

Compared against 9500GT GT220 isn't too bad. Lower power draw, higher performance, better feature set. But, chip complexity of GT216 seems to be somewhat similar to rv730, and it doesn't look too good against HD4670/4650, though at least it manages to have a somewhat lower power draw (at load - looks very similar at idle), no doubt due to 40nm manufacturing.
At suggested pricing, it's pretty bad, and it seems nvidia just hopes noone will notice GT220 isn't competitive with HD46xx, hence the silent launch.

G210 is small and low-power though again not impressive against rv710. Might not matter much though as performance isn't really a very important metric for this class of gpus.

GT240 is somewhat interesting - first nvidia product with (128bit) gddr5. Performance around that of 9800GT, with presumably lower power draw. However, quite a bit slower than the fastest cards based on good old g92b (gts250), and if pricing of GT220 is any indication, it won't really be cheaper than gts250 neither. That means it will have to fight against hd5750 likely, and that one is going to be a very very one-sided battle... I wonder what the die size of gt215 is, I suspect it could be quite similar to juniper.

[elsence]

It seems that the GT 240 will be 16ROPs and 32 TUs.

The 550MHz is very low clock for a 40nm part and based on what Nvidia could achieve with G 210 & GT 220.

I guess the design is extremely bandwidth limited with plain DDR3 (900MHz, 128bit bus) so there was no reason for an increase in core clock?

Or NV wanted also the DDR3 part to be within the PCI-Express power spec (75W)?

The performance in games of the 550MHz DDR3 part should be around 750MHz/1GHz 4670 512MB perf. or around 550MHz/800MHz G92 9600GSO/8800GS 768MB perf.

And the performance in games of the 650MHz GDDR5 part should be around 600MHz/900MHz 9800GT 512MB perf. or even around an overclocked 9800GT perf.

I hope Nvidia not to price them with the same G 210/GT220 pricing logic (The Way It's Meant To Be Priced (TWIMTBP), lol)

[Groo The Wanderer]

Or it is just too little too late. The price is waterfalled down the stack, and the real competition hasn't arrived yet. The only way that these parts will ever be competitive is if Nvidia massively subsidizes them to OEM like I have been saying. The same is true for all parts above it also, starting in a few hours......

Nvidia has no competitive parts at all right now. Zero. Their die size make it impossible to compete.

-Charlie

[DegustatoR]

Their die size make it impossible to compete.
Their die size on GT21x is much less than that of AMD's value GPUs at the same price points now. (Not that it really matters because die size alone means nothing but still.) And from the looks of it Cedar and Redwood might end up being bigger than GT21xs on the same process with the same performance.

[AnarchX]

It seems that the GT 240 will be 16ROPs and 32 TUs.

It seem to be only 8 ROPs, else it would not be slower than 9600 GT.

Their die size on GT21x is much less than that of AMD's value GPUs at the same price points now. (Not that it really matters because die size alone means nothing but still.) .
With 55nm wafers @ ~80% costs of 40nm ones and probably higher yields on 55n, I see no real advantage for:
GT216 100mm² vs RV730 140mm²
GT218 57mm² vs RV710 71mm²
(GT215s (~125mm²) die-size competitor is clearly RV740(


And from the looks of it Cedar and Redwood might end up being bigger than GT21xs on the same process with the same performance.
According to Anandtech Juniper is 166mm², so Redwood @~100mm² could be a 400SPs part and Cedar migth have the chance to be a bit faster than RV710 while be in GT218s range.

[chavvdarrr]

Their die size on GT21x is much less than that of AMD's value GPUs at the same price points now. (Not that it really matters because die size alone means nothing but still.) And from the looks of it Cedar and Redwood might end up being bigger than GT21xs on the same process with the same performance.Funny.
Few years ago someone was claiming that altho ATi has low and mid cards on smaller chips and process, still NV can compete because of using proven less expensive process. :D

[neliz]

And from the looks of it Cedar and Redwood might end up being bigger than GT21xs on the same process with the same performance.

RV810(at least 120SP) is going to be faster than RV710 (80SP), which is already 10-20% faster than GT218. So from that perspective you're getting GT216 performance for a GT218 price.

[DegustatoR]

Funny.
Few years ago someone was claiming that altho ATi has low and mid cards on smaller chips and process, still NV can compete because of using proven less expensive process. :D
Do you see any difference between "now" and "eventually"?

[Arun]

Nvidia has no competitive parts at all right now. Zero. Their die size make it impossible to compete.Charlie, you're obviously right that NVIDIA is still at a disadvantage here. Mind you, their disadvantage is much smaller than AMD's was in the R5xx timeframe or what NVIDIA had in the NV3x timeframe, so the financial consequences should not be exagerrated (of course, you could argue that NVIDIA would have been hit much harder were it not for NV33, aka the chip nobody ever heard about on the quarter-node process that barely exists but still generated the majority of NV's revenues in 2004 - still let's not get ahead of ourselves).

However, you are ignoring three things. First of all, GT216/GT2178 are more competitive in the low-end than NV's previous chips would have been for the Winter cycle and AMD hasn't refreshed that part of their portfolio yet, so NV's situation actually improved there in the last 6 months (even if it's still far from ideal to say the least)

Secondly, Fermi derivatives will come faster than you think (no, they haven't taped-out yet) and they'll be slightly more area-efficient than GF100. GT215 will probably turn out to be a fairly short-lived chip. And yes, free to come back to this post in a couple of months if I'm wrong; I'm all for accountability.

Thirdly, stopping GT200 production was a conscious move because you are right: they couldn't sell them without losing money anymore. They didn't want to repeat the fiasco that happened when RV670/G92 launched and they had too much G80 inventory. That made them lose truckloads of money. It's bad that they are in such a situation but amongst the awful strategic choices available to them, it's still (amongst) the best.

[mczak]

(GT215s (~125mm²) die-size competitor is clearly RV740

Oh that small? I thought it would be bigger (merely because GT220 is already ~100mm²). Well I guess in this case it's quite a ok chip for the die size, not really surprising it can't compete with rv740 then, maybe it'll be priced accordingly eventually.

[DegustatoR]

Oh that small? I thought it would be bigger (merely because GT220 is already ~100mm²). Well I guess in this case it's quite a ok chip for the die size, not really surprising it can't compete with rv740 then, maybe it'll be priced accordingly eventually.
G92 is able to compete with 4850, why do you think that GT215 won't be able to compete with RV740?
Another question here is what will Redwood be as that chip looks to be the one that will compete with GT215 for some time. If that's half-Juniper then that's 400 SPs/20 TUs against GT215 which should be able to reach 640 SPs/32 TUs level of RV740. Since Redwood will retain 128-bit bus (but loose GDDR5 support?) it will probably be less of a jump in die size than that between 256-bit Cypress (334?) and 128-bit Juniper (~180?). That puts it around 100-120mm^2 while having substantially less impressive specs than RV740 (which was 137mm^2) which should be the top level of GT215 performance.
I don't really see what's disasterous are there for GT21xs in light of these estimations. NV may have to cut their prices a little (GT216 should be at ~$50, GT215 at ~$80) to adjust to market conditions but that they probably will be able to do as 40G matures and TSMC's prices go down. What should be undeniable here is that all the GT21x parts will eventually cost less to produce than previous G9x parts and AMD's current RV7x0 low-cost parts. And while their timing / performance / features compared to competition may be bad I don't really see how their introduction instead of G9x GPUs and against EG 40G low-end GPUs is bad for NV in the long run.

[AnarchX]

G92 is able to compete with 4850, why do you think that GT215 won't be able to compete with RV740?

G92: 16 ROPs, 64 TMUs, 128SPs@ up to 1.8GHz
GT215: 8 ROPs (a bit improved), 32 TMUs, 96SPs(a bit improved)@ up to 1.6GHz?

And if you believe charlie NVs GDDR5 efficiency is a disaster...:lol:

[mboeller]

and 128-bit Juniper (~180?).

according to anandtech Juniper is around 166mm². So half of that would mean Redwood is <100mm²

[DegustatoR]

GT215: 8 ROPs (a bit improved)
It supports GDDR5. Why would they support GDDR5 an not have enough ROPs to utilise it?

96SPs(a bit improved)@ up to 1.6GHz?
The whole GT21x clocks thing is a bit mysterious to me right now because most of cards overclocks like hell (+30% and more than 1 GHz for ROPs in some cases). I don't understand why would NV set such low reference frequencies.

And if you believe charlie NVs GDDR5 efficiency is a disaster...:lol:
Thanks, but I tend not to believe anything that comes from a PR department.

according to anandtech Juniper is around 166mm². So half of that would mean Redwood is <100mm²
And according to Hexus authors who actually have it measured it's 12x15=180 mm^2.

[mczak]

G92 is able to compete with 4850, why do you think that GT215 won't be able to compete with RV740?
Because the hd4770 is nearly as fast as hd4850, while clearly these early benchmarks indicate gt215 isn't close to gts 250 even in its presumably fastest configuration and overclocked. Though that's only 3dmark for now, actual games remain to be seen, but I'd be surprised if things would change a lot. Clocks are suspiciously low though so maybe there's some room for improvements. Memory isn't exactly very fast neither, but since it appears to have only 8 rops faster memory might actually not help (the difference between the gddr5 version and the gddr3 version (with half the bandwidth) is already quite small).

[Jawed]

And according to Hexus authors who actually have it measured it's 12x15=180 mm^2.
Measuring the packaged die is not the same as measuring the chip. 0.5mm of packaging appears to be the norm, 11.5x14.5mm = 167mm².

Jawed

[mboeller]

And according to Hexus authors who actually have it measured it's 12x15=180 mm^2.

and I think they made a mistake.

If you compare the 5 pence (18mm diameter) with the RV790 die, the RV790 die would have a area of ~ 305 mm². Using the same measurement the RV840 die has said ~180 mm².

Therefore IMHO they didn't substracte the outer edge from the measurement when they said that the RV840 has 180 mm². Removing this outer edge the die of the RV840 is 166 mm²

So 166 mm² is imho the correct die size of the RV840.

[trinibwoy]

G92: 16 ROPs, 64 TMUs, 128SPs@ up to 1.8GHz
GT215: 8 ROPs (a bit improved), 32 TMUs, 96SPs(a bit improved)@ up to 1.6GHz?

GT216 already has 8 ROPs, what makes you think GT215 won't improve on that? Even so, I doubt it will catch G92 even with 16.

[DegustatoR]

Jawed, mboeller, does the same rule apply to GT218/216/215 die measurements? If yes then you have to adjust them by 0.5x0.5 too?

[Jawed]

Jawed, mboeller, does the same rule apply to GT218/216/215 die measurements? If yes then you have to adjust them by 0.5x0.5 too?
Should do - depends on whether the dimensions are as stated by NVidia or as measured by a reviewer including the packaging.

Jawed

[Dave Baumann]

So 166 mm² is imho the correct die size of the RV840.
Turns out that its somewhere inbetween - the 166mm² quote was actually a pre-silicon number; the final "official" size is 170mm².

[elsence]

It seem to be only 8 ROPs, else it would not be slower than 9600 GT.


With 55nm wafers @ ~80% costs of 40nm ones and probably higher yields on 55n, I see no real advantage for:
GT216 100mm² vs RV730 140mm²
GT218 57mm² vs RV710 71mm²
(GT215s (~125mm²) die-size competitor is clearly RV740(


According to Anandtech Juniper is 166mm², so Redwood @~100mm² could be a 400SPs part and Cedar migth have the chance to be a bit faster than RV710 while be in GT218s range.

When i first saw the 128bit memory controller i said 8ROPs based on previous desktop DX10 NV architectures.
But if you check the Vantage/3DMark06 scores, a 8ROPs design can't have this performance.
I mean, if the design is like you say 8ROPs/32TUs we have:

GT240 = 8ROPs / 32 TUs / 96SPs / 550MHz core / 1340MHz shaders / 900MHz GDDR5 / 57,6GB/s bandwidth

GT250 = 16ROPs / 64 TUs / 128SPs 700MHz core / 1725MHz shaders / 1000MHz GDDR3 / 64GB/s bandwidth

How is it possible a part like GT250 with 2,5X pixel fillrate with 2,5X texel fillrate and with 1,7X shading speed to be only +6% faster? (5983 vs 5613)

In the mobile space Nvidia had some designs with 8ROPs per 64bit channel, i suspect this will happen here also.

In Vantage the 550MHz GT240 (GDDR5) is around 43% faster than 600MHz 9600GT. (5613 vs 3899)

In 3DMark06 (1280X1024) the 550MHz GT240 (GDDR5) is around 2% faster than 600MHz 9600GT. (9478 vs 9224)

In 3DMark06 (1920X1200) the 550MHz GT240 (GDDR5) is around 8% slower than 600MHz 9600GT. (7629 vs 8313)

I suppose is the memory controller paired with GDDR5 (efficiency, latency, etc...)

But I don't like the Vantage scores (GT240 vs GT250) something is fishy...

If the GT216 is 486 million transistors 100mm2 die size, the GT215 should be at least 725 million transistors which means at least 137mm2 die size at 40nm (i forecast 780 million transistors with 144mm2 die size)

Regarding performance:
Redwood>RV730>GT220
Cedar>RV710>G 210

Even in the case that Redwood is a DX11 RV730 with 64bit memory controller and GDDR5 (lowest possible scenario...) it will be faster than GT220 and the die size would be 93-100mm2.

I 'll post later my predictions.

*EDIT*
Right now the 5750 512MB cost 110$.
GT 220 with 512MB GDDR3 cost 80$.
What price the GT240 DDR3 and GT240 GDDR5 will have?
At least the GT240 GDDR5 will fight against 5750 and will lose handily.

[Mintmaster]

G92 is able to compete with 4850, why do you think that GT215 won't be able to compete with RV740?

Another question here is what will Redwood be as that chip looks to be the one that will compete with GT215 for some time. If that's half-Juniper then that's 400 SPs/20 TUs against GT215 which should be able to reach 640 SPs/32 TUs level of RV740.G92 gets close to the 4850 for the same reason that it competes with the GTX 260. The latter two are cut down parts with lower clocks and/or fewer SPs. RV740 is a bit faster than a GTS250, which is substantially faster than GT215, judging by the info that has leaked out so far.

Redwood probably will compete with GT215 a bit (maybe the lower clocked ones), but only because AMD will choose higher margins rather than try to make NVidia bleed. Looking at GT216 and GT218 pricing, NVidia would rather have low sales than zero margins. AMD is in no position to turn down higher ASPs, so my guess is that Redwood and Cedar will have higher prices than we'd like.

What should be undeniable here is that all the GT21x parts will eventually cost less to produce than previous G9x parts and AMD's current RV7x0 low-cost parts. And while their timing / performance / features compared to competition may be bad I don't really see how their introduction instead of G9x GPUs and against EG 40G low-end GPUs is bad for NV in the long run.'Eventually' is a big problem, because right now ATI has successful parts that are not only mature, but are also going to be discounted when going EOL to make way for new Evergreens.

Remember that checkmark features and the Halo effect matter a lot for the value market, too.

[Mintmaster]

But I don't like the Vantage scores (GT240 vs GT250) something is fishy...Something about Vantage has always given GT2xx a marked advantage over G9x. It may be that the shaders need more registers, thus reducing the latency hiding ability and efficiency of G9x.

[elsence]

Something about Vantage has always given GT2xx a marked advantage over G9x. It may be that the shaders need more registers, thus reducing the latency hiding ability and efficiency of G9x.

Probably, but this difference is huge.

If we compare:

a GTX260 216 (1242MHz 216SP)
or a GTX275 (1404MHz 240SP)
with a GTS250 (1836MHz 128SP)

in 3DMark Vantage (Performance) scores we will see that the GT2XX has only +15% advantage per MHz*SP.

http://www.xbitlabs.com/articles/video/display/mainstream-roundup-2_13.html#sect2

In the it168.com 3DMark Vantage (Performance) tests we will see that:

a GTS250 (1725MHz 128SP) with 64GB/s bandwidth is
only 26% faster than
a GT240 DDR3 (1340MHz 96SP) with 28.8GB/s bandwidth.

So in this case the advantage is +35% per MHz*SP instead of the usual +15%.

If we compare with a GT240 GDDR5 (1340MHz 96SP) with 57.6GB/s bandwidth the advantage would be +60%.

Something is fishy...

[Groo The Wanderer]

However, you are ignoring three things. First of all, GT216/GT2178 are more competitive in the low-end than NV's previous chips would have been for the Winter cycle and AMD hasn't refreshed that part of their portfolio yet, so NV's situation actually improved there in the last 6 months (even if it's still far from ideal to say the least)

More competitive usually means less subsidies. You are right, it buys them a bit of breathing room for three months. Then again, if you look at the margins on these parts on a good day, it isn't enough to move the financial needle much. :( The segment is high volume, crap margins.

So yes, definitely improved. Above or below the profit line is still an open question.

Secondly, Fermi derivatives will come faster than you think (no, they haven't taped-out yet) and they'll be slightly more area-efficient than GF100. GT215 will probably turn out to be a fairly short-lived chip. And yes, free to come back to this post in a couple of months if I'm wrong; I'm all for accountability.

I know when they are planned, and I know why they haven't taped out yet. In fact, I think I was the first (only?) one to say it.
http://www.semiaccurate.com/2009/08/05/four-more-gt300-variants-tip/
The thing is, 'slightly' more area efficient doesn't cut it. You need to be vastly more efficient to have a chance. Then again, depending on what they cut, they could be hamstringing their own message. Watch the PR monkey dance!

Thirdly, stopping GT200 production was a conscious move because you are right: they couldn't sell them without losing money anymore. They didn't want to repeat the fiasco that happened when RV670/G92 launched and they had too much G80 inventory. That made them lose truckloads of money. It's bad that they are in such a situation but amongst the awful strategic choices available to them, it's still (amongst) the best.

I know it was a conscious move, and I wrote that, before anyone else. I also wrote up the reasons why in great detail. The thing is, with the G80, they had an imminent replacement. GF100-lite(s) are not due for minimum 3-4 months, and that is _IF_ they figure out what is wrong with Fermi the Elder.

The thing that Nvidia hasn't figured out yet is that they are making themselves look like complete fools. Right now Jen-Hsun "puppy wrangler" Huang can't look dumber. They didn't send their partners the usual EOL letters because they didn't want Wall Street to find out about their chances for Q4. I talk to a lot of them, and the spreadsheet revision has been heavy to say the least. Guess how many partners bought PCBs and components for the christmas season in anticipation of normal GPU shipments? Anyone got a use for a lot of 2xx PCBs with no chips?

Then comes the chipset fiasco. Can you look dumber than by splitting hairs over not making any more chipsets vs shutting down the chipset division? Come on, that distinction won't fly on a third grader, but sadly, many tech sites puked it back in a story. SIGH. Then there was the claim that they are stopping development because of evil Intel. Intel is _SO_ evil that they stopped AMD chipset development too!

Nvidia doesn't have a single competitive part. ATI is in control top to bottom for at least the next 3 months, likely 6. They have a cheaper part, so they can run NV margins to whatever point they feel like, and NV can't so squat. NV screwed their partners over for PR reasons, and in general, can't do anything dumber. Just wait, I will probably eat those words in a few days.

So, better position? Yeah. Which refugee camp in Dharfur would you like to call home?

-Charlie

[Mintmaster]

Probably, but this difference is huge.

If we compare:

a GTX260 216 (1242MHz 216SP)
or a GTX275 (1404MHz 240SP)
with a GTS250 (1836MHz 128SP)

in 3DMark Vantage (Performance) scores we will see that the GT2XX has only +15% advantage per MHz*SP.I doubt that's the right metric to use. 3DMark Vantage isn't solely SP limited - not by a long shot.

What I was trying to point out is that the GTX260 is 1.05x the speed of the GTS250 in 3DMark06 (using your link), but 1.37x the speed in Vantage. This jives perfectly with with the GT240 being 0.69-0.74x the speed of the GTS250 in 3DMark06 but 0.94x in Vantage.

There's nothing really fishy about it (or conversely it's been fishy ever since GT200 was released).

[Mintmaster]

Nvidia doesn't have a single competitive part. ATI is in control top to bottom for at least the next 3 months, likely 6. They have a cheaper part, so they can run NV margins to whatever point they feel like, and NV can't so squat. NV screwed their partners over for PR reasons, and in general, can't do anything dumber. Just wait, I will probably eat those words in a few days.The thing is that ATI is not going to run NV's margins into the ground. Instead, they're going to make their own margins high. NVidia did the same thing when it raped ATI in perf/mm2 back in the R5xx vs. G7x days (and, to a lesser extent, G84/G86 vs RV630/RV610).

[elsence]

I doubt that's the right metric to use. 3DMark Vantage isn't solely SP limited - not by a long shot.


I certainly agree that 3DMark Vantage isn't solely SP limited and probably this is not the right metric to use.
I used this method (which is kinda simplistic) because you said that:


Something about Vantage has always given GT2xx a marked advantage over G9x

and


It may be that the shaders need more registers, thus reducing the latency hiding ability and efficiency of G9x.

I was trying to show that probably the SPs cannot produce this advantage.
And also I was trying to show that the advantage is not the same between GT2XX and GT240.

But yes the method is kinda simplistic (but with such little data that we have what can anyone do?)

Another example is:
with GT220, Nvidia managed in relation with an oc 9500GT (lets say same MHz as GT220) with same pixel rate, with same texel rate, with +50% shaders and with same bandwidth to achieve only +30% performance advantage in Vantage per MHz.

http://www.tomshardware.co.uk/geforce-gt-220,review-31703-7.html

So why the much less powerful 550MHz GT240 GDDR5 (in relation with GTS250 per MHz, half texel, 2/3 shaders etc...) can achieve +20% speed per MHz?

I am not arguing, i just find it strange.



What I was trying to point out is that the GTX260 is 1.05x the speed of the GTS250 in 3DMark06 (using your link), but 1.37x the speed in Vantage. This jives perfectly with with the GT240 being 0.69-0.74x the speed of the GTS250 in 3DMark06 but 0.94x in Vantage.


I agree about 3DMark06, i just don't agree about Vantage.


There's nothing really fishy about it (or conversely it's been fishy ever since GT200 was released).

For me the Vantage advantage GT200 had, could be explained much more easily.
This time the Vantage advantage is much bigger, so it is not so easy to explain.

[BRiT]

XMAN26, can you keep your personal love notes to Charlie out of all public threads?

[AlexV]

If he doesn't a much longer vacation is in store for him. We're getting a bit tired of cleaning all sorts of crap, irrespective of which side of the fence generates it, so perhaps more expeditive solutions are necessary. If you want to go at one another(not you Brit, people in general) do it in RPSC or somewhere else (read another forum).

[Mintmaster]

For me the Vantage advantage GT200 had, could be explained much more easily.
This time the Vantage advantage is much bigger, so it is not so easy to explain.But I just proved to you that it's not by normalizing scores to the GTS260. Basically, the GT240 is giving scores that are 32% lower than that of GT260 in both tests.

[elsence]

But I just proved to you that it's not by normalizing scores to the GTS260. Basically, the GT240 is giving scores that are 32% lower than that of GT260 in both tests.

Yes but this imo doesn't show that everything is fine.
On the contrary it shows that something is fishy.
Maybe i am missing something?

You are saying that since the performance difference between GT240 and GTX260 is the same (around -32%) in both tests then everything is fine.
Sorry i just don't get it.
The 2 designs have different ROPs/TMUs/SPs/Bandwidth ratios, why the performance difference should be the same?
Also GTX260 has twice the speed (in specs) or more of GT240 (1,9X (Gpixels) - 2,35X (Gtexels))

Lets see the specs:

GTX260.................GT240
16,1 Gpixel............8,8 Gpixel
41,4 Gtexel...........17,6 Gtexel
112 GB/s...............57,6GB/s
216SP(1242MHz)..96SP(1340MHz)

The difference in these tests between GT240 and GTX260 is only -32%.
Why?
Let's suppose we underclock the GTX260 (-32%).

uc GTX260...........GT240
11 Gpixel.............8,8 Gpixel
28,2 Gtexel.........17,6 Gtexel
76,2 GB/s............57,6GB/s
216SP(845MHz)..96SP(1340MHz)

Why the inferior GT240 to produce the same results with the GTX260?

Sorry, i am trying to understand what you are saying, but i interpret the results differently.

I suppose soon GT240 (550MHz GDDR5) will launch.
I am trying to say that i will be very surprised If it is only -32% slower than GTX260 in Vantage because i find it strange with these specs.

[DSC]

Is the GT 220 considered a Compute 1.2 or Compute 1.1 GPU? None of the reviews even touched on this subject.

Update: Nevermind, I found the answer, it is Compute 1.2, probably one the first to be widely available.

Device 0: "GeForce GT 220"
CUDA Driver Version: 2.30
CUDA Runtime Version: 2.30
CUDA Capability Major revision number: 1
CUDA Capability Minor revision number: 2
Total amount of global memory: 1073414144 bytes
Number of multiprocessors: 6
Number of cores: 48
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 16384 bytes
Total number of registers available per block: 16384
Warp size: 32
Maximum number of threads per block: 512
Maximum sizes of each dimension of a block: 512 x 512 x 64
Maximum sizes of each dimension of a grid: 65535 x 65535 x 1
Maximum memory pitch: 262144 bytes
Texture alignment: 256 bytes
Clock rate: 1.36 GHz
Concurrent copy and execution: Yes
Run time limit on kernels: Yes
Integrated: No
Support host page-locked memory mapping: Yes
Compute mode: Default (multiple host threads can use this device simultaneously)

Specifications for Compute Capability 1.2
Support for atomic functions operating in shared memory and atomic functions operating on 64-bit words in global memory (see Section B.10);
Support for warp vote functions (see Section B.11);
The number of registers per multiprocessor is 16384;
The maximum number of active warps per multiprocessor is 32;
The maximum number of active threads per multiprocessor is 1024.

So this proves a certain ATI troll was dead wrong in claiming the GT 220 was a shrink of the G9x. It is derived from the GT200 architechture, just no FP64 ALUs. For comparison here's a G9x Compute 1.1 GPU, it has half the amount of registers.

Device 0: "GeForce 9600 GT"
CUDA Driver Version: 2.30
CUDA Runtime Version: 2.30
CUDA Capability Major revision number: 1
CUDA Capability Minor revision number: 1
Total amount of global memory: 536543232 bytes
Number of multiprocessors: 8
Number of cores: 64
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 16384 bytes
Total number of registers available per block: 8192
Warp size: 32
Maximum number of threads per block: 512
Maximum sizes of each dimension of a block: 512 x 512 x 64
Maximum sizes of each dimension of a grid: 65535 x 65535 x 1
Maximum memory pitch: 262144 bytes
Texture alignment: 256 bytes
Clock rate: 1.50 GHz
Concurrent copy and execution: Yes
Run time limit on kernels: Yes
Integrated: No
Support host page-locked memory mapping: No
Compute mode: Default (multiple host threads can use this device simultaneously)

[Mintmaster]

The 2 designs have different ROPs/TMUs/SPs/Bandwidth ratios, why the performance difference should be the same?I don't see your logic. The similarity implies that either the game workload is similar, or Vantage has parts that benefit certain ratios while 3DMark06 benefits from others, cancelling each other out. Either way, different ratios mean nothing.

For example, suppose I take 3Dmark06, and create a new benchmark called 3DMark Vintage that double the pixels, double the vertices, and double the CPU load. Every video card suffers a 50% drop in framerate. Does that contradict the fact that all these video cards have different design ratios? Of course not!
Let's suppose we underclock the GTX260 (-32%).You forgot about setup speed and CPU limitations. We already have evidence in this thread that some game tests are partially CPU limited, but more importantly, 3DMark scores do not scale down as expected with pixel count due to the high vertex workload, so all your numbers have limited applicability (since all of them only affect pixel rate).

I can't find a site with the same 3DMark06 scores for the GT240 or 9800GT, but look at resolution scaling here:
http://www.tweaktown.com/reviews/1544/gigabyte_geforce_gtx_260_graphics_card/index12.html

The scores are proportional to framerate and inversely proportional to render time, so I did some linear regression on the inverse of the scores. For example, the BFG 9800 GTX OCX has a score equal to 1/(5.16e-5 + 1.20e-11 * pixels). Try it: you'll find that that it's accurate to within five 3DMarks. The constant term (in this case 5.16e-5) is pretty similar for all cards, and indicates that 70-80% of the render time is independent of pixel count. Only the 1.20e-11 figure will go down with more shaders/BW/ROPs/TMUs, thus they can only slightly increase your 3DMark score.

Now, the vga.it168.com review seems to have more strenuous settings, and the weaker GT240 seems to spend about 45% of its 3DMark06 render time limited by anything pixel related at 1920x1200 (again, that's BW, ROPs, TMU, or ALUs), and I'll wildly guess that maybe 15% more is vertex shader limited. If you double the speed at which pixels and shaders are processed/output, then you'll only take off 30% of the render time, and you'll only get a 42% higher score.

I can't find any Vantage data at different resolutions without changing the settings (e.g. AA, AF), but a similar situation holds. If a GTX260 spends 55% of the render time limited by CPU or setup, then a GT240 (which has roughly half the BW/ROPs/ALUs/TMUs) will take 1.45x as long to render the scene, not 2x as long. Or, put another way, the GT240 is 31% slower.

Lets see the specs:

GTX260.................GT240
16,1 Gpixel............8,8 Gpixel
41,4 Gtexel...........17,6 Gtexel
112 GB/s...............57,6GB/s
216SP(1242MHz)..96SP(1340MHz)(Hence my "roughly half" estimate during pixel limited parts of the workload.)

[AnarchX]

Full review of GT 240 @ IT168:
http://translate.google.ch/translate?u=http%3A%2F%2Fvga.it168.com%2Fa2009%2F1 019%2F763%2F000000763107.shtml&sl=zh-CN&tl=en&hl=de&ie=UTF-8

~ 9600 GT performance level on reference clocks.

[Blazkowicz]

the naming is a shame (a gddr3 version should be called "GT230"), and that cooler looks small.

give me a gddr-5 model with a dual slot cooler and I won't complain.
I could actually make use of that card if it's cheap enough. It doesn't use much power, apparently less than a radeon 4770 (which compares roughly to a 9600GT).

It would make a great energy efficient card (I don't mind gaming at 1024x768 high IQ anyway)

[Psycho]

It doesn't use much power, apparently less than a radeon 4770 (which compares roughly to a 9600GT).

Ehh.. I see no 4770 numbers there but from other reviews (where system power in furmark is 30-40w below 4850) it looks like it's only 5-10 watt less in both load and idle, while the 4770 would be around 40% faster (the 4850 is generally 40-50% faster here, and the 4770 is only a few percent behind it).

[mczak]

the naming is a shame (a gddr3 version should be called "GT230"), and that cooler looks small.

If those benchmarks we saw are any indication the performance difference between gddr5 and "sddr3" (I wonder if this actually really is gddr3 and not just ddr3) isn't too big. But I won't disagree that there still should be a different name.


give me a gddr-5 model with a dual slot cooler and I won't complain.

They tested the gddr5 version. I think you won't see dual slot coolers with this class of chips. Generally unnecessary and just driving costs up. Doesn't mean you couldn't fit one yourself, of course.


I could actually make use of that card if it's cheap enough. It doesn't use much power, apparently less than a radeon 4770 (which compares roughly to a 9600GT).

It would make a great energy efficient card (I don't mind gaming at 1024x768 high IQ anyway)
Well, it only seems to use a little less power than the hd4770, while losing quite badly in performance. Still, I think it's an ok card. No external power connector, a pcb looking very simple (and hence cheap). So as long as it's actually sold cheaper than the hd4770 it should find its place. At least it beats the HD4670 (would have been disaster if not), though I suspect the ddr3 version will have to fight pretty hard to beat redwood-based cards (and those should be cheaper, use even less power, have more features etc. etc.).

[Blazkowicz]

Gigabyte makes a dual slot GT220, which looks good except it's half the specs at maybe 80% the price. shader clocks are pretty high, the GT240 probably has headroom.

right now I'm running an overclocked 8400GS and was quite surprised, it's faster than one might think.

[trinibwoy]

Full review of GT 240 @ IT168:
http://translate.google.ch/translate?u=http%3A%2F%2Fvga.it168.com%2Fa2009%2F1 019%2F763%2F000000763107.shtml&sl=zh-CN&tl=en&hl=de&ie=UTF-8

~ 9600 GT performance level on reference clocks.

Hmmm, that's impressive efficiency wise given the lower clocks and unit counts. Too bad the price is too high and absolute performance is too low.

[mczak]

Hmmm, that's impressive efficiency wise given the lower clocks and unit counts. Too bad the price is too high and absolute performance is too low.
What lower unit counts? It's got 96 (granted clocked slightly slower) instead of 64 shader units, and the same amount of texture units. It has (maybe) only half the rops and lower core clock too, but I'd say performance is right there where you'd expect it compared to 9600GT. And certainly on a perf per area metric it's less than stellar (a shrinked g94 should be smaller in theory).

[trinibwoy]

What lower unit counts? It's got 96 (granted clocked slightly slower) instead of 64 shader units, and the same amount of texture units. It has (maybe) only half the rops and lower core clock too, but I'd say performance is right there where you'd expect it compared to 9600GT. And certainly on a perf per area metric it's less than stellar (a shrinked g94 should be smaller in theory).

Yeah you're right, brain fart. Don't know why I thought it only had 16 TMUs and not 32. Nevermind then, not so impressive at all. Not sure it makes sense to compare its die size to a best case G94 shrink though....

[Blazkowicz]

poor nvidia lacks a whole GPU here.

[Tchock]

EDIT: Oh this is GT215. My bad.

~12*12 = 144

excluding 0.5mm for package, 11.5^2= 132mm^2.

[mczak]

Actually thinking about this, the (g)ddr3 versions might already have to fight pretty hard with HD4670 if it loses 20% or so performance with the slower ram. That wouldn't be so good, rv730 has very similar size but using 55nm...

[AnarchX]

Halloween launch party for a "franken card" with mix of GT21x GPUs?
http://nvnews.net/vbulletin/showthread.php?t=140298
Maybe GT215 SLI/GT200 with GT216 for PhysX on one PCB?

edit:
EVGA teaser: http://www.evga.com/articles/00512/
http://www.evga.com/articles/00512/images/footer.jpg

The shape of the mysterious card matches to a Single-Slot GTX 295: http://www.evga.com/PRODUCTS/IMAGES/GALLERY/017-P3-1295-AR_XL_4.jpg

[Alexko]

EDIT: Oh this is GT215. My bad.

~12*12 = 144

excluding 0.5mm for package, 11.5^2= 132mm^2.

Is it just me, or is that really poor 40nm scaling?

[AnarchX]

Is it just me, or is that really poor 40nm scaling?

Do not forget that GT216 is 486 million transitors, while G96b was 314 mio.
If the scaling is around the same for G94 to GT215, it should be a bit over 700 million.
This should be over 5.5 mio transistor per square-mm, not good as Evergreen but not bad for GT2xx design.

[neliz]

It's a "meh"

http://forums.slizone.com/index.php?showtopic=39989

The F34R Channel: Nvidia & EVGA to release GTX275 (graphics) + GTS250 (PhysX) on 1 card, single PCB on 10-30-09
From our understanding, NVIO2 chip is connected to GTX 275, while GTS 250 cannot display an image at all. The sole purpose of GTS 250 is to be used for PhysX effects - during regular day in Windows, the G92b GPU sits in idle mode and consumers moderate amount of power. The chip is only powered up when PhysX gets into overdrive. Do note that the data mentioned above is our estimate."

[trinibwoy]

Wow, Nvidia really is going buck-wild with this physics thing. Pretty optimistic to sell hardware that's only useful for a single good game though :eek:

[fellix]

Frankenstein's GeForce. :lol:

[Vincent]

Frankenstein's GeForce. :lol:


Two GPUs + two distinct memory configurations ??

[Vincent]

It's a "meh"

http://forums.slizone.com/index.php?showtopic=39989


G92 for free. :grin:

[compres]

Excess G92 stock? ;)

[neliz]

Excess G92 stock? ;)

Now you have to wonder if these are the GPU's that failed to qualify as Keychain stock :p

[rpg.314]

It's a "meh"

http://forums.slizone.com/index.php?showtopic=39989

well this is an out and out physx card...., I am not sure how many excess g92's they can get rid of this way but it allows them to charge premium for it.

And they can finally beat the cuda/physx balloon drum harder than ever. with physx on, they'll finally be able to beat a 5870. :)

[mczak]

And they can finally beat the cuda/physx balloon drum harder than ever. with physx on, they'll finally be able to beat a 5870. :)
You don't need separate g92 chip for that, the single-threaded cpu implementation of physx makes sure it's slow enough that a single gt200 doing both physx and graphics is always faster :-).
That said, if true that's really a strange card. I think the price premium such a combo is going to have won't really be justifiable however.

[Arun]

Now you have to wonder if these are the GPU's that failed to qualify as Keychain stock :pThose G98s achieve an amazing 0% yield afaik, so I don't think even they would have been cynical enough to sell it as a PhysX accelerator :)

Is it possible that it is GT200 inventory they've got too much of rather than G92 inventory? They're still going to (have to) sell some G92s for some time, so it'd be strange if they felt they had to get rid of them in a hurry.

[Tchock]

@Arun GT240 desktop (OEM) was originally supposed to be a G92, but now it's all 40nm, so I would assume that G92 is getting EOLed, GT200 stocks are the one to clear (but not that much either) and the 40nm parts live on.

[ChrisRay]

Its just a halloween promotion guys...

[SiliconAbyss]

Its just a halloween promotion guys...
Meaning, what?

Is this a limited time product? A limited volume product?

[neliz]

Its just a halloween promotion guys...

Which was completely free to design?
Someone invested money in this.. this.. horror...

[ChrisRay]

Which was completely free to design?
Someone invested money in this.. this.. horror...

Someone invested money into designing full GT200 295 cards as well with 1 gig of memory and all the ROPS. So what? I think you guys overestimate how much something like this takes to build.

It's a neat little gadget. Some people may even like the concept. It just is really wierd how you all are making a big deal out of this.

[Davros]

It just is really wierd how you all are making a big deal out of this.

Arnt you supposed to be talking up nv products not downplaying them ;)

[ChrisRay]

This is an EVGA made board. :)

[AnarchX]

4Gamer.net says there are rumors about a 40nm GTX 260 in industries since September: http://www.4gamer.net/games/095/G009528/20091105001/

[Cookie Monster]

4Gamer.net says there are rumors about a 40nm GTX 260 in industries since September: http://www.4gamer.net/games/095/G009528/20091105001/

Could this be the supposed GT212? Specs say 216SP, 256bit (possibly for GDDR5?) and DX10.1. Looks like a GTX260~275 filler depending on its clock speed.

That roadmap seems to make sense, using the GT21x as a filler til the rest of the DX11 Fermi derivatives arrive.

[CJ]

Could this be the supposed GT212? Specs say 216SP, 256bit (possibly for GDDR5?) and DX10.1.

GT212 had 384 SPs.

[Cookie Monster]

GT212 had 384 SPs.

Was that ever confirmed? Well I dont know, but this supposed 216SP 40nm GPU looks like a good replacement for the mid range market.

[CJ]

Was that ever confirmed? Well I dont know, but this supposed 216SP 40nm GPU looks like a good replacement for the mid range market.

Officially confirmed... no.... since these companies never talk about unannounced/unreleased products. But the correct specs have been spotted on roadmaps before and even posted on some sites like Hardware Infos (http://www.hardware-infos.com/news.php?news=2629). And before you ask... yes those specs on that site are correct.

[trinibwoy]

Heh, things would've been mighty different if it had actually made it to market then. Though Q2 2009 seems mighty aggressive, especially for Nvidia.

[no-X]

GT212 appeared in the roadmap earlier, than GT200b. It seems, that GT200b was a back-up plan.

[LordEC911]

GT212 appeared in the roadmap earlier, than GT200b. It seems, that GT200b was a back-up plan.

Jensen doesn't approve of backup plans. :wink:

[no-X]

Surprisingly, the old roadmap (w/o GT200b/GTX285) is still in the first post:

http://img398.imageshack.us/img398/4585/nvidiaroadmapgt212gt206wz6.jpg

[Kaotik]

But wasn't deemed later that GT206 = GT200b?

[no-X]

Maybe CJ could tell us more(?)

[Arun]

GT206 was a G98 replacement that got canned.

[Groo The Wanderer]

Officially confirmed... no.... since these companies never talk about unannounced/unreleased products. But the correct specs have been spotted on roadmaps before and even posted on some sites like Hardware Infos (http://www.hardware-infos.com/news.php?news=2629). And before you ask... yes those specs on that site are correct.

Was this from the same source that gave him the 512b memory, <480mm^2, 2.xb transistor GT300 info, or the other mole that gave him the 1200 shader 'R870', and the amazing R870X2 with faster memory than the vanilla R870?

I can't honestly say which one was funnier, him posting those specs, or Theo 'confirming' them from 'his sources'. Both had me laughing quite hard.

-Chalrie

[DegustatoR]

GT212 appeared in the roadmap earlier, than GT200b. It seems, that GT200b was a back-up plan.
It seems that this roadmap was wrong to begin with.

[neliz]

It seems that this roadmap was wrong to begin with.

Other people above you already confirmed it's existence.

[no-X]

Local PR confirmed existence of GT212, but he also siad, that GT212 was canceled shortly after this roadmap went public.

[DegustatoR]

Other people above you already confirmed it's existence.
Where exactly? When Arun said that GT206 was a G98 replacement while this roadmap shows it as a GTX260 replacement?

[neliz]

Where exactly? When Arun said that GT206 was a G98 replacement while this roadmap shows it as a GTX260 replacement?

ici (http://forum.beyond3d.com/showpost.php?p=1355867&postcount=884)

[DegustatoR]

ici (http://forum.beyond3d.com/showpost.php?p=1355867&postcount=884)
I'm not saying that GT212 never existed. I'm just saying that Elsa's particular roadmap was wrong.

[neliz]

I'm not saying that GT212 never existed. I'm just saying that Elsa's particular roadmap was wrong.

Yes, it wouldn't have launched in Q1 but a couple of months later.

[no-X]

The roadmap was old and outdated when it appeared publicly. But that doesn't prove it was wrong since the day 1.

[Vincent]

GT212 had 384 SPs.



384 SP ??? :grin:

So, Is it the first Geforce version derived from the Fermi architecture ( No DP)?

or just another DX10.1 GT200 with extra SPs ?

[Ailuros]

Whatever GT212 exactly was if you look at this vaporware's GT21x@40nm grandchildren today I doubt it was worth the hussle at all.

[trinibwoy]

Why is that? Do they underperform their specs?

[Ailuros]

Why is that? Do they underperform their specs?

http://www.flickr.com/photos/42786968@N05/sets/72157622520411613/

The results are so boring that it's not even worth mentioning against a 9600GT; and yes there are obviously differences in spefications between the two, but it's still not a product that would turn any heads. Probably some OEMs for the fancy 10.1 sticker but that's about it. If the GT215 would be clocked higher it would of course make a difference.

[Jawed]

http://www.techreport.com/discussions.x/17979

The GeForce GT 240 has a 40-nm DirectX 10.1 GPU, just like the GeForce GT 220 and G210, but it also features 96 stream processors, a 550MHz core clock speed, a 1340MHz shader speed, a 128-bit memory interface, and either 512MB or 1GB of memory. That memory can be either 850MHz GDDR5 (i.e. 3400MT/s, or total memory bandwidth of 54.4GB/s) or slower, 1000MHz DDR3. The card draws 70W under load.

[trinibwoy]

Wow, Nvidia finally has GDDR5 parts in retail....thought I'd never see the day.

[AnarchX]

According to PCGH (http://www.pcgameshardware.de/aid,699574/Test-Geforce-GT-240-Nvidias-schnellste-DirectX-101-Grafikkarte/Grafikkarte/Test/) 727 mio. transistors.

[neliz]

Wow, Nvidia finally has GDDR5 parts in retail....thought I'd never see the day.

Lol... no .. they ANNOUNCED a GDDR 5 part ;)
Seen how all the other GDDR5 parts are on backorder for the next few weeks it looks like a December launch to me (correct me if I'm wrong.)

(it also doesn't help that it's priced higher than the HD4850)

edit: I am wrong, european price searches turn up nada but it is available in the US... but still $100 for a 9600GT?

[Jawed]

http://techpowerup.com/reviews/MSI/GeForce_GT_240/1.html

SLI is done over PCI Express as there's no SLI connectors. Still have no idea how big the die is. Or the fillrates.

Jawed

[Cookie Monster]

Guru3d review (http://www.guru3d.com/article/msi-geforce-gt-240-review-test/)

[Speccy]

SLI is done over PCI Express as there's no SLI connectors.

Have you actually seen SLI benchmarks?

http://www.tomshardware.co.uk/geforce-gt-240,review-31731-2.html

Hoping that SLI might be supported over the PCI Express bus, we tried running our two test samples together. However, the driver panel wouldn't show us the option to enable SLI. Confused, we asked Nvidia for a bit of clarification. The company let us know that the GeForce GT 240 does not support SLI, and that Nvidia "typically hasn’t supported SLI for sub-$99 products, as users typically upgrade instead of buying a second card."

[XMAN26]

Have you actually seen SLI benchmarks?

http://www.tomshardware.co.uk/geforce-gt-240,review-31731-2.html

Which is why we can find SLi connectors on 9500GTs?! go figure.

[Tchock]

Oooh gosh, better have everything in hardware next time. nVidia seems to be on a locking spree these days.

[Jawed]

Have you actually seen SLI benchmarks?
No I hadn't. The article I linked said so and I left it at that :oops:

Jawed

[mczak]

This is the only review so far which I've seen which had both GDDR5 (1700Mhz) and DDR3 (only 800Mhz) memory cards: http://www.pcgameshardware.de/aid,699574/Test-Geforce-GT-240-Nvidias-schnellste-DirectX-101-Grafikkarte/Grafikkarte/Test/

The results are imho (but we already knew this) a bit disappointing. The gddr5 version struggles to keep up with the GT9600 (guess mostly due to half the rops?), while the ddr3 version has a hard time beating the HD4670.
Why is that chip so big btw? Die size identical to rv740 (no way it keeps up with that...), transistor count almost the same as g92 while having way less functional units (half the rops and texture units, 3/4 the alus, and no unless GT200 it can't claim wasted transistors for DP). Oh right it got DX10.1...
Though power/performance is quite good, at least for now it can rightfully claim the title "fastest card without pcie power connector", since its power consumption is about the same as HD4670 and (at least the GDDR5 version) is definitely around 20% or so faster on average.
SLI connector missing a cost cutting measure? Despite that and oem's cutting further corners (silly noisy fan without fan control) priced too high.

[neliz]

Can you imagine that it was targeted as a G92 replacement some time back?
I think Degustator promoted it a couple of pages back:

G92 is able to compete with 4850, why do you think that GT215 won't be able to compete with RV740?

It would be all fine and dandy if it was actually running at the clocks he's talking about (1Ghz) and not the clocks it's running on now (550 and 650 for the to be anounced OC models)

[Tchock]

Oh, and can someone who has the card TRY to overclock the memory over 1Ghz (2/4Ghz)?

I'd like to see if something blows up or if the card accidentallys something (most likely the whole thing)

[DegustatoR]

Still have no idea how big the die is.
139 mm^2 according to PCGH.

[Jawed]

139 mm^2 according to PCGH.
And we still have to guesstimate 0.5mm per dimension of packaging, I presume, which would result in about 128mm².

At a sane 1680x1050:

http://techpowerup.com/reviews/MSI/GeForce_GT_240/30.html

RV740 is 143% performance for 107% die size on the same process. 150%+ performance at higher resolutions.

Jawed

[no-X]

deleted

[Mintmaster]

Why is that chip so big btw? Die size identical to rv740 (no way it keeps up with that...), transistor count almost the same as g92 while having way less functional units (half the rops and texture units, 3/4 the alus, and no unless GT200 it can't claim wasted transistors for DP).Because it comes from the GT200 architecture. It's a step back from G9x in performance density.

The groundwork for GT2xx was probably laid while ATI still had R600. They wanted to add features for CUDA (DP, larger register file, maybe more), and felt they had room. When RV670 came out, they probably thought it was a decent improvement over R600 in competitiveness, but this was the best that ATI could optimize and it was still a decent step back from G9x. They probably also expected higher clock speeds.

If GT215 was fabbed at 65nm, it probably would be around 250 mm2. Now consider how big GT200 is compared to G92 and the limited peformance gain you get. Using these two pieces of infomation, it's not surprising that GT215 is a bit below 9600GT performance. Transistor numbers are largely PR, so don't get to caught up in that.

[Blazkowicz]

It's nvidia's 4670. only it needs to be priced like GT220 ddr3 currently is, plus we need a dual slot or passive model rather than that OEM garbage.
Note that frequency looks a bit conservative and power usage is low overall, so it might be o/cable a bit. (I don't consider o/c on the big guzzling cards)


whoops, I didn't see the techpowerup review, with a dual slot MSI modern. Weird comments about the noise. Sure, there's no fan control, but I don't get how 35dB is that noisy. e.g. the 4850 is 10dB less when idle but 5dB more under load.
isn't that a constant slight hum, versus a fan controlled card spinning up when gaming or even watching fullscreen youtube? I downgraded to an o/c 8400GS to get rid of that annoyance. playing the waiting game, and the 8400GS is actually brilliant at running older games.

[Jawed]

Because it comes from the GT200 architecture. It's a step back from G9x in performance density.
Does GT215 have double precision support?

The groundwork for GT2xx was probably laid while ATI still had R600. They wanted to add features for CUDA (DP, larger register file, maybe more), and felt they had room.

http://www.techreport.com/articles.x/14934/5

Another enhancement in GT200 is the doubling of the size of the register file for each SM. The aim here is, by adding a more on-chip storage, to allow more complex shaders to run without overflowing into memory. Nvidia cites improvements of 35% in 3DMark Vantage's parallax occlusion mapping test, 6% in GPU cloth, 5% in Perlin noise, and 15% overall with Vantage's Extreme presets due to the larger register file.

Graphics benefits from increased register file (much like the gain from increased ALU:TEX) - CUDA gains are riding on the back of that. It was silly small to start with in G80. It's increasing again in Fermi, that might appease the CUDA crowd.

GT200 also gained improvements in texturing throughput - presumably that costs some area, but we have no idea what.

The D3D10.1 changes (doubling the per vertex attributes, cubemap features, "fp32 everywhere", gather4 etc.) will also have cost an uncomfortable amount, I expect.

If GT215 was fabbed at 65nm, it probably would be around 250 mm2. Now consider how big GT200 is compared to G92 and the limited peformance gain you get. Using these two pieces of infomation, it's not surprising that GT215 is a bit below 9600GT performance. Transistor numbers are largely PR, so don't get to caught up in that.
I think the real reason GT215 is so poor is that it seems it only has 8 ROPs (same as HD4670). The very high Z rate compensates somewhat as it's double the per ROP rate of ATI, though 47% faster in absolute theoretical terms, and it has 70% more bandwidth. It should have 16 ROPs, to compete against RV740, but that would make it significantly larger.

In my view NVidia has made a GDDR3 budget chip to replace 9500GT, low power and cool for OEM systems and it happens to have GDDR5 as a testbed for later chips.

Jawed

[DSC]

GT21x GPUs also have PureVideo HD VP4, which adds bitstream decoding of VC-1, MPEG-2 and MPEG4 ASP in addition to H.264, compared to VP2 which only support H.264 bitstream decoding and partial accelerated decoding of VC-1 and MPEG-2.

Also, HD audio processor is built-in with 8 channel LPCM support. All previous GeForce 8/9 and the GT200 GPUs didn't have the HD audio processor with 8 channel LPCM support.

Both of these features will take up slightly more transistors & die space than the GeForce 8/9 design.

The only disappointing things are the low shader clock speed, 1340MHz vs 1625MHz on the 9600 GT and half the ROPs instead of the expected 16 ROPs. If it had say, 1500MHz shader clock or higher & 16 ROPs then it should easily beat the 9600 GT. GT 220 has pretty large headroom, Gigabyte sells a 720Mhz core 1566MHz shader clock overclock version instead of the 625MHz core 1360MHz shader clock stock GT 220.

[mczak]

GT21x GPUs also have PureVideo HD VP4, which adds bitstream decoding of VC-1, MPEG-2 and MPEG4 ASP in addition to H.264, compared to VP2 which only support H.264 bitstream decoding and partial accelerated decoding of VC-1 and MPEG-2.

Also, HD audio processor is built-in with 8 channel LPCM support. All previous GeForce 8/9 and the GT200 GPUs didn't have the HD audio processor with 8 channel LPCM support.

Both of these features will take up slightly more transistors & die space than the GeForce 8/9 design.
This functionality is found also in the lowest end products, I doubt it takes up a significant amount of area. Sure it will use more area than in the old cards but how much? 2 mm² or what?
The only disappointing things are the low shader clock speed, 1340MHz vs 1625MHz on the 9600 GT and half the ROPs instead of the expected 16 ROPs. If it had say, 1500MHz shader clock or higher & 16 ROPs then it should easily beat the 9600 GT. GT 220 has pretty large headroom, Gigabyte sells a 720Mhz core 1566MHz shader clock overclock version instead of the 625MHz core 1360MHz shader clock stock GT 220.
Yes, lower clocks is certainly some trend we're seeing with GT200 architecture cards. That said, they are really low here, especially if you consider you can get the same chip as mobile, with lower power draw and actually slightly higher clocks (gts260m)!
With clocks around 650/1550 it should (even with 8 rops) be able to beat GT9600 quite solidly, though it would need the additional power connector and still couldn't touch HD 4770, so I guess nvidia went with the OEM-friendly version instead. Though I'm wondering why they didn't come up with two cards, I think the large quantities will be with ddr3 versions anyway, so could have as well sold as cheap, no-pcie power connector ddr3 version and higher clocked gddr5 version (with a different name - but they just love to sell different cards with the same name and same cards with different names I know...).

And actually I take the "fastest card without pcie power connector" back. That title would have to go to the 9800GT Green Edition, I suppose.

[trinibwoy]

Doesn't make too bad of a showing here vs 9600GT - http://www.computerbase.de/artikel/hardware/grafikkarten/2009/test_nvidia_geforce_gt_240/18/

[Tchock]

That title would have to go to the 4850 Green Edition, I suppose.

Fixed.
Chinese manufacturers are scarier than you thought.

Wow @ the 4670 actually- 128bit and GDDR3 and on a bigger process, it doesn't get bulldozed (75% and below). Let's see how the future mobile Junipers do- if they scale down and such with clocks and voltage.

For now, this is quite an impressive mobile chip for the GTX 250M SKU I have to say.

[mczak]

Fixed.
Chinese manufacturers are scarier than you thought.

Well yeah but that's not an official part. No doubt rv740/juniper parts would also fit into a power envelope not requiring additional power with some different clocks, but these don't exist neither.


Doesn't make too bad of a showing here vs 9600GT

Depends on how you look at it. It's a way more complex chip, with more shader alus (which are supposedly faster too), similar memory bandwidth. But still 10% slower (on average) with AA/AF at least (and 9600GT is using older driver there so might even gain a little with newer one). Though maybe it'll at least end up cheaper - currently it's not.

[trinibwoy]

Depends on how you look at it. It's a way more complex chip, with more shader alus (which are supposedly faster too), similar memory bandwidth. But still 10% slower (on average) with AA/AF at least (and 9600GT is using older driver there so might even gain a little with newer one). Though maybe it'll at least end up cheaper - currently it's not.

Yeah it does depend on how you look at it. For example how does it fare compared to other 8-ROP cards? In fillrate limited cases the 9600gt has a major advantage. I agree it does seem complex for its specs - maybe DX10.1 was expensive.

[DSC]

Fudzilla's review sample card has very good overclocking, 650MHz core, 1590 shader and 2200MHz memory, practially maxed out. All this without the need for external power.

http://www.fudzilla.com/content/view/16492/1/1/5/

Techpowerup's sample got up to 670MHz core, 1670MHz shader and 2190MHz memory. Again without external power.

http://www.techpowerup.com/reviews/Palit/GeForce_GT_240_Sonic/33.html

The other Techpowerup review sample managed 620MHz core, 1610MHz shader and 1960MHz memory.

http://www.techpowerup.com/reviews/MSI/GeForce_GT_240/33.html

So there is headroom. Perhaps Nvidia is being conservative with the core and shader clock speed, since the 40nm TSMC process has issues and of course, being a more OEM friendly SKU.

[gkar1]

Geforce GT 240, Geforce GT 220 and Geforce G210 were launched on the market to keep Nvidia's name in the press loop...We doubt that these GPUs will maintain any high market share in the long run.

Thanks Fudo, but we knew all of this months ago. What I want info on is Fermi mainstream and low end variants, assuming they even exist at this point in time.

[mczak]

Fudzilla's review sample card has very good overclocking, 650MHz core, 1590 shader and 2200MHz memory, practially maxed out. All this without the need for external power.
It is somewhat likely though it'll violate PCIE spec and draw more power than 75W at these settings. No doubt this will still work, but it probably wouldn't be a very good idea for stock settings...

[CarstenS]

Doesn't make too bad of a showing here vs 9600GT - http://www.computerbase.de/artikel/hardware/grafikkarten/2009/test_nvidia_geforce_gt_240/18/

Given that it's appearing some 18 months after that on a new process node with about 50% more transistors and a heftier price tag, a "too bad of a showing" is a bit less than one should be allowed to expect.

[Mindfury]

NVIDIA GeForce 310

http://h10010.www1.hp.com/wwpc/images/emea/VG885AA_400x400.jpg

http://h10010.www1.hp.com/wwpc/uk/en/sm/WF06c/A1-329290-64268-348724-348724-4015767-4015770.html

[trinibwoy]

Given that it's appearing some 18 months after that on a new process node with about 50% more transistors and a heftier price tag, a "too bad of a showing" is a bit less than one should be allowed to expect.

That was obviously in relation to previous, more horrible showings not to any reasonable expectations. :wink:

[neliz]

NVIDIA GeForce 310

http://h10010.www1.hp.com/wwpc/images/emea/VG885AA_400x400.jpg

http://h10010.www1.hp.com/wwpc/uk/en/sm/WF06c/A1-329290-64268-348724-348724-4015767-4015770.html

Yay! GT218 rebranding arrives! .. How many months did that take since its launch?

Specs:

CUDA Cores 16
Graphics Clock (MHz) 589 MHz
Processor Clock (MHz) 1402 MHz
Memory Clock (MHz) 500
Standard Memory Config 512 MB DDR2
Memory Interface Width 64-bit
Memory Bandwidth (GB/sec) 8.0
Feature Support:
NVIDIA PureVideo® Technology* yes
NVIDIA CUDA™ Technology yes
Microsoft DirectX 10.1
OpenGL 3.1
Bus Support PCI-E 2.0
Certified for Windows Vista yes
Display Support:
Maximum Digital Resolution 2560x1600
Maximum VGA Resolution 2048x1536
Standard Display Connectors DVI
VGA
DisplayPort
Multi Monitor yes
HDCP* yes
HDMI* Via dongle (DVI-HDMI or DP-HDMI)
Audio Input for HDMI Internal
Standard Graphics Card Dimensions:
Height 2.731 inches
Length 6.60 inches
Width Single-slot
Thermal and Power Specs:
Maximum GPU Temperature (in C) 105 C
Maximum Graphics Card Power (W) 30.5 W
Minimum System Power Requirement (W) 300 W

Specs:
CUDA Cores 16
Graphics Clock (MHz) 589 MHz
Processor Clock (MHz) 1402 MHz
Memory Specs:
Memory Clock (MHz) 500
Standard Memory Config 512 MB DDR2
Memory Interface Width 64-bit
Memory Bandwidth (GB/sec) 8.0
Feature Support:
NVIDIA PureVideo® Technology* yes
NVIDIA CUDA™ Technology yes
Microsoft DirectX 10.1
OpenGL 3.1
Bus Support PCI-E 2.0
Certified for Windows Vista yes
Display Support:
Maximum Digital Resolution 2560x1600
Maximum VGA Resolution 2048x1536
Standard Display Connectors DVI
VGA
DisplayPort
Multi Monitor yes
HDCP* yes
HDMI* Via dongle (DVI-HDMI or DP-HDMI)
Audio Input for HDMI Internal
Standard Graphics Card Dimensions:
Height 2.731 inches
Length 6.60 inches
Width Single-slot
Thermal and Power Specs:
Maximum GPU Temperature (in C) 105 C
Maximum Graphics Card Power (W) 30.5 W
Minimum System Power Requirement (W) 300 W

[Blazkowicz]

I have to buy one. 310 is the biggest three-digit-number for nvidia cards, so it has to be better than a 220 or a 260?

[neliz]

I have to buy one. 310 is the biggest three-digit-number for nvidia cards, so it has to be better than a 220 or a 260?

310 is like 15 more than a 295! .. wait.. I made that joke before..

How soon before someone tries to run anything DX11 related on this card? "But the websidez told me it waz the new GF300 sseries! man! Firmy!"

[Alexko]

Yay! GT218 rebranding arrives! .. How many months did that take since its launch?

Specs:


Specs:

Jesus Christ, not again! And what the hell are they going to do when they release their Fermi derivatives? Call them GT4xx?

[trinibwoy]

Yep

[thatdude90210]

Jesus Christ, not again! And what the hell are they going to do when they release their Fermi derivatives? Call them GT4xx?
And renamed GT5xx a month later.

[CarstenS]

Yay! GT218 rebranding arrives! .. How many months did that take since its launch?

Don't be so mean! After all, they slightly changed the zoom on the products' picture on their website ;-)

[neliz]

Don't be so mean! After all, they slightly changed the zoom on the products' picture on their website ;-)

I'm pretty sure that you have to sign up for the "At Least The People Who Print New Boxes Are Happy" program.

[CJ]

310 is OEM only afaik. So you won't see this as a retailcard. But yeah, their naming scheme is just... confusing.

[CarstenS]

So was G210 - before it consumed the confuser… err…

[fellix]

210 vs. 310 (http://img689.imageshack.us/img689/8996/oldnewnvidia.gif)

:D

[Ailuros]

Every PC needs good rebranding :twisted:

[mczak]

310 is OEM only afaik. So you won't see this as a retailcard. But yeah, their naming scheme is just... confusing.
Really, last nvidia graphic card series where naming made any sense at all was GF8xxx - and I'm talking about the G8x based parts, not G9x. Since then it's nothing but confusing (GT240 vs GTS240? You think they are almost the same?), just marketing ploys (GTX260M etc.), and rebranding just to have some new products (GTS250, GF310, ...).

[Blazkowicz]

some rebranding are not so evil, "9800GTX+" sucks as a name.
off course it was all fucked up since the moment they named the 8800GT as such rather than 9800GT straight away.

[Alexko]

some rebranding are not so evil, "9800GTX+" sucks as a name.
off course it was all fucked up since the moment they named the 8800GT as such rather than 9800GT straight away.

Yeah, I never understood why the 9800 GTX+ wasn't named 9900 GTX or something like that.

But Nvidia should really drop all the suffixes and prefixes and stick to numbers.

[rpg.314]

And I thought nv was done was done with it's rebranding shenanigans. Sigh......

[neliz]

And I thought nv was done was done with it's rebranding shenanigans. Sigh......

If a Certain Someone is to be believed, we might see G92 again as well. And I don't think the 275 Co-Op side entry counts for that.

[rpg.314]

If a Certain Someone is to be believed, we might see G92 again as well. And I don't think the 275 Co-Op side entry counts for that.

This rebranding crap has gone on for too long now. I just hope redwood and cedar land here pronto and clear out all this relabeled/rebranded BS.

[trinibwoy]

I still don't get why rebranding for the OEM market makes sense. The OEM's themselves obviously won't be fooled and most consumers won't even know the difference so why bother?

[Alexko]

I still don't get why rebranding for the OEM market makes sense. The OEM's themselves obviously won't be fooled and most consumers won't even know the difference so why bother?

Consumers will see machines with "GeForce 210" and machines with "GeForce 310". Guess which one they'll pick...

Some of them might even know that Nvidia's current high-end is the GTX 285/295, so they'll figure that this is newer.

[Blazkowicz]

consumers will get it as part of their PC, and if game at all, they will be amazed by its speed and quality (it's an order of magnitude upgrade other their previous 9200SE or Intel graphics)

[trinibwoy]

Consumers will see machines with "GeForce 210" and machines with "GeForce 310". Guess which one they'll pick...

Why does Nvidia care if they sell a 210 or 310? It's the same thing to them.

Some of them might even know that Nvidia's current high-end is the GTX 285/295, so they'll figure that this is newer.

Why would they? I didn't assume my mid-range receiver was better than the high-end last generation version because the model number was higher. People aren't morons. It's fair to assume it's better than the last generation version in the equivalent price range though - an assumption that doesn't hold up here.

[neliz]

Why would they? I didn't assume my mid-range receiver was better than the high-end last generation version because the model number was higher. People aren't morons. It's fair to assume it's better than the last generation version in the equivalent price range though - an assumption that doesn't hold up here.

I had a discussion recently with someone that was looking for a PCI FX5200. his old card had died but he wanted the same card back and not some of that new ATI crap.. he had $50 to spend.

But in general, yeah .. the "price" and "looks" of the card would show everyone this isn't in the same class.

[Groo The Wanderer]

Yeah, I never understood why the 9800 GTX+ wasn't named 9900 GTX or something like that.

But Nvidia should really drop all the suffixes and prefixes and stick to numbers.

Nvidia should come out with cards that justify the new numbers, then drop the suffixes and prefixes. Those only come about when a company is floundering without product to sell.

If you don't have the product, spin. Anyone remember probably the most egregious example, the ATI X800XTX-PE? :) That coined the term 'Press Edition', I don't think any were actually sold. at all.

-Charlie

[neliz]

If you don't have the product, spin. Anyone remember probably the most egregious example, the ATI X800XTX-PE? :) That coined the term 'Press Edition', I don't think any were actually sold. at all.

-Charlie

In a mainstreet in Taipei, they were selling them in aluminium suitcases with plexi-glass top. As for me, I just bought the Pro-Vivo and flashed it, saved me $150.

[FrameBuffer]

Nvidia should come out with cards that justify the new numbers, then drop the suffixes and prefixes. Those only come about when a company is floundering without product to sell.

If you don't have the product, spin. Anyone remember probably the most egregious example, the ATI X800XTX-PE? :) That coined the term 'Press Edition', I don't think any were actually sold. at all.

-Charlie

ahh yes the fabled X800XTX- Phantom Edition.. belongs in the annals of graphics history along side the 6800 GTX Ultra Extreme Edition, 7800 GTX 512MB and Asus Mars...

[Silent_Buddha]

ahh yes the fabled X800XTX- Phantom Edition.. belongs in the annals of graphics history along side the 6800 GTX Ultra Extreme Edition, 7800 GTX 512MB and Asus Mars...

Yeah, except it actually sold a fair amount and at a cheaper price. I still remember 7800 GTX 512's going for 1000 USD. I think Newegg even got 5 of them to sell.

Anyways, I'm a proud owner of a X800XTX-PE. :D Could have sworn it was just X800XT-PE. I'll have to see if I still have the box...

Well, I'm glad at least one graphics company has stopped rebranding (so far). I'll be happy when the other company stops it also.

I'm getting tired of explaining the differences (or lack of differences) to customers and friends when they ask me about the newest Nvidia cards and whether they should replace their current ones. And it turns out their currents cards are virtually the same as the new ones, just with the older name. /sigh...

And then they think I'm an idiot because it MUST be different since the name is different. Then I have to spend an hour or two explaining and sometimes arguing with them... UGH.

Regards,
SB

[FrameBuffer]

Yeah, except it actually sold a fair amount and at a cheaper price. I still remember 7800 GTX 512's going for 1000 USD. I think Newegg even got 5 of them to sell.

Anyways, I'm a proud owner of a X800XTX-PE. :D Could have sworn it was just X800XT-PE. I'll have to see if I still have the box...

Well, I'm glad at least one graphics company has stopped rebranding (so far). I'll be happy when the other company stops it also.

I'm getting tired of explaining the differences (or lack of differences) to customers and friends when they ask me about the newest Nvidia cards and whether they should replace their current ones. And it turns out their currents cards are virtually the same as the new ones, just with the older name. /sigh...

And then they think I'm an idiot because it MUST be different since the name is different. Then I have to spend an hour or two explaining and sometimes arguing with them... UGH.

Regards,
SB

Yeah for some reason nV is very capable of having a strangle hold on the $1000 video card market where as ATI can't sell the market's top performing graphics card for less than the GTX 280 intro'd at w/o catching flack.. then again nv also seems to have cornered the market on renaming where as ATi's last rename (that i can recall) was the RV630.. some 3 generations ago. I think someone correctly pointed out that most re-branding/re-naming seems to go hand in hand when a company is floundering .. just as ATI was during HD2XXX and now with nV and the whole G92->b->210->310 ->410 ??

[mczak]

then again nv also seems to have cornered the market on renaming where as ATi's last rename (that i can recall) was the RV630...
Did these cards (hd 4250 and friends) actually ever appear in the market? Can't remember I've actually seen one for sale...

[Groo The Wanderer]

I'm getting tired of explaining the differences (or lack of differences) to customers and friends when they ask me about the newest Nvidia cards and whether they should replace their current ones. And it turns out their currents cards are virtually the same as the new ones, just with the older name. /sigh...

And then they think I'm an idiot because it MUST be different since the name is different. Then I have to spend an hour or two explaining and sometimes arguing with them... UGH.

Just do what I do, push the parts that are named correctly. Any vendor that screws with it's customers like that does not deserved to be promoted.

-Charlie

[thatdude90210]

Just curious. Did Nvidia shed some engineers. Did they lose some to Intel... or cashed out their stock options and retired? The company used to execute like clockwork. These last couple of years, all they got out was the GTX200/b. GTX200 derivatives are a year late & a dollar short, Fermi is late, and derivatives are never even mentioned.

[Alexko]

Why does Nvidia care if they sell a 210 or 310? It's the same thing to them.

The issue isn't whether they sell a 210 or a 310, but rather how many they sell. And I suspect introducing a newer model boosts sales, otherwise they wouldn't do it. And they've been doing that sort of thing for a while now, so it probably works.


Why would they? I didn't assume my mid-range receiver was better than the high-end last generation version because the model number was higher. People aren't morons. It's fair to assume it's better than the last generation version in the equivalent price range though - an assumption that doesn't hold up here.

I said newer, that doesn't necessarily imply faster, but new is good, right? And yes, people are morons, but more importantly they're not experts, and 300 > 200 so they may not always think beyond that. They might also know that the GTX 260/80s and Radeon HD 4850/70s were released around the same time, and therefore figure that the GeForce 310 is somehow equivalent to the Radeon HD 5000s.

[trinibwoy]

The issue isn't whether they sell a 210 or a 310, but rather how many they sell. And I suspect introducing a newer model boosts sales, otherwise they wouldn't do it. And they've been doing that sort of thing for a while now, so it probably works.

They sell chips to OEM's and AIB's, not to consumers. Why would they buy more of something because it has a new sticker on it?

I said newer, that doesn't necessarily imply faster, but new is good, right? And yes, people are morons, but more importantly they're not experts, and 300 > 200 so they may not always think beyond that. They might also know that the GTX 260/80s and Radeon HD 4850/70s were released around the same time, and therefore figure that the GeForce 310 is somehow equivalent to the Radeon HD 5000s.

You must assume that people behave more stupidly when buying OEM PCs than they do when buying other electronic devices.

[Alexko]

They sell chips to OEM's and AIB's, not to consumers. Why would they buy more of something because it has a new sticker on it?

Because the new sticker will help OEMs sell more computers to consumers.



You must assume that people behave more stupidly when buying OEM PCs than they do when buying other electronic devices.

I think on average people behave fairly stupidly, but mostly they don't know the ins and outs of computer hardware, and NV's constant renaming makes it even more confusing. It's easier with other devices: when you're buying an iPod, it's pretty clear that the 80GB version can store twice as much music as the 40GB one. If you don't really know what you're looking at, a GeForce 9400 1024MB might look a lot better than a 512MB one, but we both know that's not the case. I think Silent_Buddha's post just confirms that.

[Silus]

I said newer, that doesn't necessarily imply faster, but new is good, right? And yes, people are morons, but more importantly they're not experts, and 300 > 200 so they may not always think beyond that. They might also know that the GTX 260/80s and Radeon HD 4850/70s were released around the same time, and therefore figure that the GeForce 310 is somehow equivalent to the Radeon HD 5000s.

That's quite a fallacy and still it's thrown around constantly when discussing this "issue"...

People take price over anything to assume something is better than something else i.e. if A is more expensive than B, then it means A is better than B. That's just common sense for those not "in the know", so no one in that group would think the GeForce 310 is somehow equivalent to the HD 5000s...

And are people morons, just because they don't follow tech news or look at hardware specs thoroughly ?

[mczak]

People take price over anything to assume something is better than something else i.e. if A is more expensive than B, then it means A is better than B. That's just common sense for those not "in the know", so no one in that group would think the GeForce 310 is somehow equivalent to the HD 5000s...

That's quite true, though for OEM parts you often can't really see the price of the individual components. I guess the expectation people have is that a pc costing twice as much as another one is just faster at everything - of course it might just have much faster cpu but slower gpu indeed. In that case though it won't matter at all what the graphic card is named.
GT310 though probably sets a new record for rebranding in shortest time, GT210 has barely been on the market (1 month or so?)...

[Alexko]

That's quite a fallacy and still it's thrown around constantly when discussing this "issue"...

People take price over anything to assume something is better than something else i.e. if A is more expensive than B, then it means A is better than B. That's just common sense for those not "in the know", so no one in that group would think the GeForce 310 is somehow equivalent to the HD 5000s...

And are people morons, just because they don't follow tech news or look at hardware specs thoroughly ?

Not equivalent to the HD 5800s and 5700s, but to the low-end 5000s, they could very well assume that.

And no, they're not morons because they don't follow tech news, it's completely unrelated.

[ShaidarHaran]

Not equivalent to the HD 5800s and 5700s, but to the low-end 5000s, they could very well assume that.

And no, they're not morons because they don't follow tech news, it's completely unrelated.

But certainly there are those out there that believe higher number = better/more performance as was stated. After all, that is the naming scheme, albeit minus the all-important caveat that performance starts with the 2nd digit and the 1st digit is just the generation.

[Silent_Buddha]

But certainly there are those out there that believe higher number = better/more performance as was stated. After all, that is the naming scheme, albeit minus the all-important caveat that performance starts with the 2nd digit and the 1st digit is just the generation.

Except Nvidia doesn't seem particularly concerned with the first number denoting generation. :( But they expect their customers to think that, easy way to bilk them...

Granted this was general practice for both Nvidia and AMD in the past, but at least AMD has moved past that (and hopefully won't go back). Now just need Nvidia to hop aboard.

Regards,
SB

[ShaidarHaran]

Except Nvidia doesn't seem particularly concerned with the first number denoting generation. :( But they expect their customers to think that, easy way to bilk them...

Granted this was general practice for both Nvidia and AMD in the past, but at least AMD has moved past that (and hopefully won't go back). Now just need Nvidia to hop aboard.

Regards,
SB

Renaming certainly complicates things, but even the performance of these parts is accurately reflected by the model number.

[Psycho]

Renaming certainly complicates things, but even the performance of these parts is accurately reflected by the model number.

In other words we should expect the 300 series to be roughly as fast as the similar 200 series? :wink:

[ShaidarHaran]

:lol: doubt that very much. Performance of products based upon model number alone is only meant to be compared within the same generation.

[Silent_Buddha]

Well if 310 does equal 210 when it comes out, Nvidia is basically saying performance of 3xx generation is the same as 2xx generation, indirectly. :D

Renaming works both ways... :P And either way, just wish everyone would stop doing it.

Regards,
SB

[AnarchX]

GT216 (GT 220?) is GeForce 300 series part, too:
NVIDIA GeForce 315 1GB PCIe x16 Card Form Factor ATX Full Height
Display max resolution Analog VGA: 2048x1536x32bpp@85Hz
DVI: 2560 x 1600 x32bpp @60Hz
Input/Output connectors DVI-I +VGA
Board configuration Specification Description
Graphics Chip D10M2-20 (GT216-200-A2)
Core clock 625 MHz
Memory clock 790 MHz
Frame buffer 1GB DDR3, 128-bit
Core power 52 W (Max board power)
http://h18000.www1.hp.com/products/quickspecs/13494_div/13494_div.HTML

[neliz]

Remember the Charlie's nv Roadmap?

http://www.semiaccurate.com/static/uploads/2009/09_september/NV_season_roadmap.jpg

[Silent_Buddha]

Gah, so basically GT230 and below = GT330 and below... Well, except that GT330 has a little bit more memory....

Woo, way to position your Fermi generation (if it indeed carries the 3xx nomenclature). Way to go out of your way to say that 3xx generation performance is exactly the same as 2xx performance...

/blarf...

Well at least Nvidia finally matched AMD with Dx10.1... After AMD released Dx11 based cards...

Even the whole blunder with Nv30 didn't leave Nvidia this far behind the competition technology wise.

Regards,
SB

[Albuquerque]

Even the whole blunder with Nv30 didn't leave Nvidia this far behind the competition technology wise.

I was actually just thinking this earlier. I didn't see the NV30 series as entirely horrible (as I know some do); I had a friend with an FX5600 and he seemed to like the card -- and it performed well for what he wanted.

Still, the NV30 line was new technology, and NV was still on technology par at the time -- albeit, not performance par ;) And G80 was even early to the DX10 party, and to put icing on the cake, it ended up being one of NV's biggest defining performance parts. How long did that architecture allow them to sit in the #1 spot?

But now they're behind the technology curve; more so than likely any other point in their history. At least the NV30 was a product that they could produce and deliver -- they're not even done spinning the silicon if the rumors are to be believed. I don't think NV30 was the holocaust that everyone says, but I also believe that this is turning out worse than the NV30 debacle for a multitude of reasons.

[thatdude90210]

Anyone else take that JHH's recent interview where he said something to the effect that (paraphrasing) "leaders don't use backup plans" to mean that they are already on their backup plan?

[CarstenS]

. At least the NV30 was a product that they could produce and deliver -- they're not even done spinning the silicon if the rumors are to be believed. I don't think NV30 was the holocaust that everyone says, but I also believe that this is turning out worse than the NV30 debacle for a multitude of reasons.

With NV30, rumor mills started churning what super duper kind of chip it was roughly from October 2002, first samples and early performance tests where to be had from January 2003 and the card hit the fan (or was it something other that hit it?) in february/march 2003.

Now it's +7.

[no-X]

CarstenS: First (truthful) rumours were published even earlier. I remember news about nVidia GPU, which will be the first one supporting super-fast next-gen DDR memory. It was sometimes during winter 2001/02 :-)

[Groo The Wanderer]

Gah, so basically GT230 and below = GT330 and below... Well, except that GT330 has a little bit more memory....

Woo, way to position your Fermi generation (if it indeed carries the 3xx nomenclature). Way to go out of your way to say that 3xx generation performance is exactly the same as 2xx performance...

/blarf...

Well at least Nvidia finally matched AMD with Dx10.1... After AMD released Dx11 based cards...

Even the whole blunder with Nv30 didn't leave Nvidia this far behind the competition technology wise.

Regards,
SB

That is _SO_ unfair! Nvidia is only matching ATI on DX10.1 in the lowest three tiers, G215, G216 and G218. The rest are DX10, after all, why do you need 10.1? Go back and check the quotes from NV execs if you aren't sure.

Also, NV launched DX10.1 cards several weeks before ATI launched DX11. They were out, as OEM only parts, in syphilitic trickle quantities, for such a large number of weeks that you could count it on one hand. That'll show those Canadian commies what real engineering is about!

-Charlie

[LordEC911]

Anyone else take that JHH's recent interview where he said something to the effect that (paraphrasing) "leaders don't use backup plans" to mean that they are already on their backup plan?
That's an interesting take on the quote.
I took it as the usual stockholder/analyst BS from JHH, similar to the "nothing is wrong" and "we are going to open a can of whoop(s)ass."

[trinibwoy]

I think he was trying to communicate that they're fully committed to whatever strategy they choose. But that's a poor choice of words as it implies they're too narrow-minded to change course if things aren't working out.

[thatdude90210]

I saw it as a way to deflect possible criticism. Their halo product was late, whole dx11 line was even more late, people might start asking why there was no "plan b" where they add dx 11 functions to existing chips (I have no idea how difficult that may or may not be, afterall, they did do enough work to add dx10.1) to avoid missing a significant MS launch. So he's saying "pfff, leaders don't do backup plans." Basically like Peewee Herman falling off his bike saying "I meant to do that."

[neliz]

I saw it as a way to deflect possible criticism. Their halo product was late, whole dx11 line was even more late, people might start asking why there was no "plan b" where they add dx 11 functions to existing chips (I have no idea how difficult that may or may not be, afterall, they did do enough work to add dx10.1) to avoid missing a significant MS launch. So he's saying "pfff, leaders don't do backup plans." Basically like Peewee Herman falling off his bike saying "I meant to do that."

He said "easy as (insert explicit noun here) in the same interview..

I’m not cynical about the market but If you give them the same thing over and over again, they’ll get tired. Look at iPod- there’s iPod fatigue. If you were to have an iPod, well, yeah- who doesn’t? If you have a Zune HD now, everybody wants to look at it- people need a new thing. It is possible to build a new thing. Thats the reason why we didn’t wanna take the GT200 and add DX11. Its possible. Its as easy as peach.

He's both lambasting his own renaming shenanigans and saying they could have a DX11 compliant GT200 (GT21x surely?) out in a matter of months.
Heck, according to him they could have DX11 out last year. Now I wonder why it took so many moons to get dx10.1 parts out. Surely, he's playing the readers for fools.


Dear reader,
In winter no DirectX11
have fun

[Silus]

Remember the Charlie's nv Roadmap?

Yeah, it's still as ridiculous as it was when he posted the first time :)

[sethk]

I wonder what constitutes 'Spring 2010' in that roadmap, and why divide it into 'Holiday 2009' and then 'Spring 2010' , what about the gap in between? Spring is generally March-April-May. (It doesn't actually align on the start and of the months, but I'm rounding it out.) If by mid-May Nvidia is still trying to push the GTX280 as their high end part, I'll be shocked.