ATI RV740 review/preview

[Kaotik]

Regardless of the whole 9800/GTS issue, nVidia really went too far with the GTX260M and GTX280M, G92b based mobile chips

 

[ninelven]

This is actually something Nvidia is trying to address with their new naming schemes. Currently it is unfriendly to the consumer. To go just by numbering. However nvidia wants to emphasize further on the extension rather than the numbers this time.

Think

"GTX Performance"
"GTS Midrange"
"GS Mainstream"

If Nvidia sincerely wants to help the consumer, this would be better:

GTX - Performance
GT - Midrange
G - Mainstream

...just saying.

 

[codedivine]

Uhh .. somewhat back to RV740 .. does anyone wish to speculate if its brining new compute features?
Also I wonder if Rv740 will have double precision (which is the only compute feature missing from radeon 4670 compared to 48x0).

 

[ChrisRay]

If Nvidia sincerely wants to help the consumer, this would be better:

GTX - Performance
GT - Midrange
G - Mainstream

...just saying.

Actually its

GTX
GTS
GT

Honestly. What difference does the extension name make? As long as its clearly labeled in its intended performance segment? Ti/MX worked fine. I dont see why it wouldn't this way.

 

[CarstenS]

Fuad has "pics"

I've actually seen (and fotographed) said notebook too. To me, it seems like there was noch HD 4860 inside - because if it were, why would also the memory be identified wrong? It also had a fairly recent driver from 13th of february customized for Asus...

 

[Silent_Buddha]

Actually its

GTX
GTS
GT

Honestly. What difference does the extension name make? As long as its clearly labeled in its intended performance segment? Ti/MX worked fine. I dont see why it wouldn't this way.

If that's ALL it was, it wouldn't be quite as bad, although I'm still glad I no longer have to keep track of those things with one vendor.

But then you have (+)/(216)/etc...

Granted you need SOME way to differentiate different cards with the exact same name, but at some point it just gets silly in order to claim one-ups-manship in the fastest video card war.

And as far as I'm concerned, whether it's ATI or Nvidia doing it. Rebranding the EXACT same chip to multiple generations is just trying to take advantage of consumers that don't know better.

And G92b branded as mobile GTX 260 and mobile GTX 280? Really? That just takes the cake. I'm honestly completely flabbergasted at this point that Nvidia is so desperate to con customers that they would go to this extent.

It's disgusting when ATI does it and it's disgusting when Nvidia does it.

Regards,
SB

 

[Mariner]

If thats true then its true for every generation of GPU from both AMD and NV. The Radeon 4550 is not better than the Radeon 3870 and yet it would appear to be from its name alone. Do you therefore have a problem with the Radeon 4550?

I feel this renaming of cards by the big IHVs is an actively disingenuous attempt to sell more cards to the ignorant.

It doesn't matter who does it (although in this case we were obviously talking about NVidia's latest renaming shenanigans). It's still dishonest.

 

[ninelven]

I dont see why it wouldn't this way.

I wouldn't expect you to. However, for the vast majority of people on this planet, GTS vs GTX is much harder to distinguish than Ti vs MX (and I knew plenty who couldn't get that right).

 

[trinibwoy]

AMD hands down has the most concise and straightforward model numbering scheme. Nvidia's prefixes are still clunkier in comparison. However, there's no way to reliably inform customers about differences between generations without them at least reading the spec sheets and/or reviews. The same goes for cars, tv's, digital cameras.....

 

[pjbliverpool]

I think that CJ was talking about improvements in general... So HD4600 did bring improvements (other than strictly performance-related) over HD3800: smaller, lower consumption, cheaper, easier to squeeze into notebooks, etc...
GTS250 brings no improvement over 9800GTX+, other than its name.

The GTS 250 in its 1GB form also brings a lower price and lower power consumption over the 9800GTX+.

And were the 4600 brought those advantages over the 3800 at the cost of performance. The 250 has no such cost associated with in in comparison to the 9800+

But all thats beside the point anyway. The point is, what advantages would a 250 based on GT2xx have had over the current one? And aside from supporting a later version of CUDA which is meaningless to the vast majority of consumers, I see none. And yet it would have had to cost more to recoup its R&D. Hence a net loss for the consumer.

 

[pjbliverpool]

Regardless of the whole 9800/GTS issue, nVidia really went too far with the GTX260M and GTX280M, G92b based mobile chips

Yeah I agree on that one. If they had wanted to name them in line with the desktop range then they could quite easily have been called GTS 240M and GTS250M. The current names are blatently misleading, especially considering its much harder to understand the performance and specs of mobile GPU's due to a relative lack of reviews.

 

[eastmen]

Originally Posted by pjbliverpool View Post
If thats true then its true for every generation of GPU from both AMD and NV. The Radeon 4550 is not better than the Radeon 3870 and yet it would appear to be from its name alone. Do you therefore have a problem with the Radeon 4550?

ati's hardware is pretty easy to understand

The first number denotes generation of chip and the second two numbers denote its performance within that generation of chip.

So no the 4550 is not faster than the 3870. However it is from the new generation , consumes much less power and is faster than the 3550. At the same time the 4870 is much faster than the 3870. Its not as confusing as whats going on with nvidia as they are naming old generation chips with new generation names. So mabye the gt 250 is faster than the gt 240 , but if the gt 240 is a new generation of chip it may have other featuers over the gt 250 aside from speed that it lacks inlcluding but not limited to power savings.

 

[DegustatoR]

Yeah I agree on that one. If they had wanted to name them in line with the desktop range then they could quite easily have been called GTS 240M and GTS250M. The current names are blatently misleading, especially considering its much harder to understand the performance and specs of mobile GPU's due to a relative lack of reviews.

AMD is doing exactly the same with RV740 and 4830/4860 Mobility Radeons.

HD 4860: 640 SP, 650 MHz, 832 GFlops, Up to 4.0 Gbps 128 bit GDDR5
HD 4850: 800 SP, 625 MHz, 1000 GFlops, 2.0 Gbps 256 bit GDDR3

So how exactly is the first one faster than the second?
Both companies are the same. Both are using strange and alogical naming schemes. You all have to use your brain to figure it out every time, sorry.

 

[Arty]

AMD is doing exactly the same with RV740 and 4830/4860 Mobility Radeons.

HD 4860: 640 SP, 650 MHz, 832 GFlops, Up to 4.0 Gbps 128 bit GDDR5
HD 4850: 800 SP, 625 MHz, 1000 GFlops, 2.0 Gbps 256 bit GDDR3

So how exactly is the first one faster than the second?
Both companies are the same. Both are using strange and alogical naming schemes. You all have to use your brain to figure it out every time, sorry.

Maybe you missed CJ's post earlier. The Mobility 4850 is clocked lower, hence RV740 was slotted as 4860.

AMD is not naming RV670 as Mobility Radeon 4800, IMO you are mixing two completely different logics.

 

[DegustatoR]

Maybe you missed CJ's post earlier. The Mobility 4850 is clocked lower, hence RV740 was slotted as 4860.

GTX 280M is faster than 9800M GTX so NV is doing exactly the same.

AMD is not naming RV670 as Mobility Radeon 4800, IMO you are mixing two completely different logics.

RV670 is the same as RV770 only slower. Again -- exactly the same as between RV740 and RV770.

I sense a disturbing lack of logic in this thread.

 

[pjbliverpool]

ati's hardware is pretty easy to understand

The first number denotes generation of chip and the second two numbers denote its performance within that generation of chip.

Its only easy to understand if you already know what each of the numbers represent.

Anyone ignorant enough to think a mid range GTS 250 is faster than a top end 9800GTX+ is highly unlikely to know what each number represents in the ATI naming scheme.

All they will see is one number being bigger than the other.

So no the 4550 is not faster than the 3870. However it is from the new generation , consumes much less power and is faster than the 3550.

Does it consume much less power than the 3550? I wouldn't mind seeing some evidence to support that.

I'm not sure of what relevance it is though. The GTS 250 is much faster than NV's equivilent from the last generation as well - the 9600GT. It probably consumes more power but tbh, i'm confused as to exactly when lower power consumption became a "feature" of a new GPU architecture. In general, power consumption goes up, or at least remains roughly steady as GPU generations progress. And consumers (especially the ones that don't frequent forums like this) rarely care about, or even bother to check the power consumption of a new GPU, except to make sure it will run with their power supply. What matters is price, performance, and technical capabilties. usually in that order.

 

[Arty]

GTX 280M is faster than 9800M GTX so NV is doing exactly the same.

*sigh*

What stopped them from naming it GTS 250M? Clearly its vastly under powered to the GTX280.

RV670 is the same as RV770 only slower. Again -- exactly the same as between RV740 and RV770.

I sense a disturbing lack of logic in this thread.

Maybe I should have worded it a bit better. Did AMD offer a higher clocked RV670 as Mobility Radeon 4800? No, its not equivalent and your comparison simply falls flat.

Of all the renaming shenagians, G92b ending up as GTX 280M takes the cake. There is no excuse defending that.

 

[A.L.M.]

The GTS 250 in its 1GB form also brings a lower price and lower power consumption over the 9800GTX+.

And were the 4600 brought those advantages over the 3800 at the cost of performance. The 250 has no such cost associated with in in comparison to the 9800+

You said it right. "In its 1GB form". The new board design is required only on the 1GB model, while on the 512MB the new design is not required.
No cost of performance? NVidia partners are just about to launch their lower-clocked 1GB cards, which could lower GTS 250 performance below the level of 9800GTX+.

But all thats beside the point anyway. The point is, what advantages would a 250 based on GT2xx have had over the current one? And aside from supporting a later version of CUDA which is meaningless to the vast majority of consumers, I see none. And yet it would have had to cost more to recoup its R&D. Hence a net loss for the consumer.

The point is why rename a card if it remains still the same?

GTX 280M is faster than 9800M GTX so NV is doing exactly the same.

You're mixing up things... ATi would have named RV740 Mobility HD4770, but since 40nm led to higher clocks, it ended up to be faster than the underclocked RV770. So to avoid confusion, they call Mobility HD4860 what should have been named HD4770. Anyway, the difference of a Mobility HD4860 from a HD4850 is not so big. This cannot be said about the GTX280 M, which is much slower than a desktop GTX280. It's utterly misleading.
GT130M isn't any faster than 9700GT M. SO why rename it? The same applies to the 9800GS M, which became the GTS 160 M.

RV670 is the same as RV770 only slower. Again -- exactly the same as between RV740 and RV770.

I sense a disturbing lack of logic in this thread.

Hmm, what the heck...

 

[pjbliverpool]

You said it right. "In its 1GB form". The new board design is required only on the 1GB model, while on the 512MB the new design is not required.

True, but its still bringing last gen top end performance into current gen mid range performance while maintaining the same feature set as the lastest generation of high end GPU's. I'm not sure what else people are expecting from this chip. I'll say it again though, what advantage do you think the 250 would have had over the current incarnation if it had been based on a GT2xx? In what way would the consumer have benefited?

No cost of performance? NVidia partners are just about to launch their lower-clocked 1GB cards, which could lower GTS 250 performance below the level of 9800GTX+.

Yep, I definatly disagree with the bait and switch tactic. There is no justififcation for that IMO. However the exact same tactic could have been pulled with a GT2xx based 250. It has absolutely nothing to do with the rename.

And besides, even a slight decrease in performance has no baring on the argument. Its quite usual for a midrange part of the current generation to perform worse than the top end card of the previous generation. For both IHV's.

The point is why rename a card if it remains still the same?

1) To simplify the naming scheme for for the consumer. As I said in a previous post, GTS 240, GTS 250, GTX 260, GTX 280 is much clearer than 9800GT, 9800GTX+, GTX 260, GTX 280.

2) And yes, to boost sales, why not? The 9800GTX+ is a perfect candicate for the GTS250 position. It lacks no features, it lacks no performance, and its cheap. If consumers are irrationaly shying away from buying a 9xxx card simply because it sounds "last generation" then I see no problem with "re-educating" them with this name change. Afterall, for all intents and purposes, the 9800GTX+ IS a midrange GT2xx.

 

[Dave Baumann]

RV670 is the same as RV770 only slower. Again -- exactly the same as between RV740 and RV770.

How do you figure that? There are considerable architectural differences between the two and that results on functional differences as well.