The NEXT LAST R600 Rumours & Speculation Thread [Archive] - Page 8

nicolasb

10-Apr-2007, 10:17

Fudzilla latest:

http://www.fudzilla.com/index.php?option=com_content&task=view&id=462&Itemid=1

Confirms that official name will be Radeon X2900.

Also:

We can also confirm that the R600 will have that audio part, but that won't be a classic sound card, that will be video related somehow. We will cover this one later, when we gain some more details about it.

And mentions that R600 has the integrated UVD (Universal Video Decoder) that has previously been talked about in connection with RV6xx.

Continuing here:

http://www.fudzilla.com/index.php?option=com_content&task=view&id=463&Itemid=1

R600 won't exactly compete with the X-Fi audio processors from Creative labs. It will have an audio processing part, but this one will concentrate on High definition video.

Radeon X2900 has a UVD, Dedicated Universal Video Decoder part of the chip, dedicated for Video only. The decoder supports Blu-ray and HD-DVD, but it really doesn't care about the format as the UVD processes H.264 content regardless of the format.

R600 has a Xilleon video processor inside and this video decoder is an all-in-one integrated solution.

Integrated stand alone sound supports HDCP and HDMI and works well under Vista. It supports full HD sound. As far as we know today, R600 sound will be related to Video but there is a chance that you will have an R600 audio driver that will play your MP3's. We will try to get this for you.So, some confirmation that R600 will have the same audio capabilities as RV6xx. This also suggests that the audio capability is limited to the passing through of (say) BluRay or HD-DVD disc soundtracks via HDMI - excellent news for HTPC users, but of little interest to anyone else.

If (like me) you're about to put together a system that is intended both for gaming and also for HTPC use, it sounds as though R600 might still be worth waiting for!

nicolasb

10-Apr-2007, 13:09

Fudo's on a roll:

http://www.fudzilla.com/index.php?option=com_content&task=view&id=464&Itemid=1

R600, Radeon X2900 series will have some cool video features. We already told you a bit here. When we went deeper in the stuff we've seen we learned that R600 can process two independent video streams.

The big R600 chip has a video part of chip called resource schedule that can process two independent video streams, even in High definition.

Radeon X2900 will be able to play your HD content as s first stream while you will be able to watch a second stream as a picture in picture feature. You saw this in most high quality LCD and High end TVs. This will be cool and definitely indicates that R600 unified Shaders can do more than graphics, they can take care of two HD streams.Actually it may not indicate anything of the sort. As I understand it the UVD section of the chip is separate from the shaders. (But I could well be wrong).

Kaotik

10-Apr-2007, 14:11

Fudo's on a roll:

http://www.fudzilla.com/index.php?option=com_content&task=view&id=464&Itemid=1

Actually it may not indicate anything of the sort. As I understand it the UVD section of the chip is separate from the shaders. (But I could well be wrong).

UVD is indeed separate from the "shadercore", the way I understood it, it's basicly Xilleon integrated to the same silicon as the rest of the core.

NocturnDragon

10-Apr-2007, 16:04

Fudo's on a roll:

http://www.fudzilla.com/index.php?option=com_content&task=view&id=464&Itemid=1

Actually it may not indicate anything of the sort. As I understand it the UVD section of the chip is separate from the shaders. (But I could well be wrong).

Couldn't it be that when you try to watch two different video HD streams UVD takes care of the first and the sencond one is processed using shaders?

compres

10-Apr-2007, 19:09

Couldn't it be that when you try to watch two different video HD streams UVD takes care of the first and the sencond one is processed using shaders?

Thats exactly what I thought. Is it not the way it is done on current solutions from ATI?

Kaotik

10-Apr-2007, 19:18

Thats exactly what I thought. Is it not the way it is done on current solutions from ATI?

In the current solutions shadercore is working on any videostream you might throw at it as far as I know, the xilleon based part on it handles other things than actually processing the video

Sound_Card

10-Apr-2007, 19:56

By my friend AlexxisF1 at R3D.

http://i49.photobucket.com/albums/f279/allexxisf1/agenafx.jpg

http://i49.photobucket.com/albums/f279/allexxisf1/r600_name.png

AMD is aiming for the benchmark numbers and plans to let the media advertise for them. What better way to rule them all than by having a 8 core K10 system equiped with HT3, HTX, PCIe 2.0, and 4 R600's in quad fire? Can you say 2 1/2 tera flops in a box?

Geeforcer

10-Apr-2007, 20:01

I was under the impression that Agena was not due till sometine in Q308.

Russell

10-Apr-2007, 20:09

I was under the impression that Agena was not due till sometine in Q308.

That's when it's due. But since Agena FX is really just a pair of Barcelona's, which will be released sooner, it's reasonable that you can show them off a bit early.

3dilettante

10-Apr-2007, 20:13

Kaotik

10-Apr-2007, 20:14

I was under the impression that Agena was not due till sometine in Q308.

Q307 you mean? Anyway, probably is, but using it now isn't any different from for example Intel releasing benchmark of Core 2 several months before actually putting those for sale

Razor1

10-Apr-2007, 20:58

I think you guys mean Q3 2007.

If AMD is showing Agena anything, these would be engineering samples (or an overclocked Barcelona where they won't open the case or show the CPUID).

Basically, it would be "here's a chip you won't see again for six months winning a few benchmark comparisons that won't be valid in six months".

Yeah and marketing is pointless for a chip that won't be released for a good six months, in those 6 months there are going to be countless press reviews for the r600, with no mention of AMD CPU's, this could backfire, because the PR of showcasing the r600 on Agena on editor's day will have reviews that don't give the same shining outlook. Not to mention editor's day material will be under NDA until the reviews hit, so there will be almost no impact on the general consumer.

w0mbat

10-Apr-2007, 21:10

I was under the impression that Agena was not due till sometine in Q308.

Agena |= AgenaFX

Kaotik

10-Apr-2007, 21:18

trinibwoy

10-Apr-2007, 21:20

Yeah and marketing is pointless for a chip that won't be released for a good six months, in those 6 months there are going to be countless press reviews for the r600, with no mention of AMD CPU's, this could backfire, because the PR of showcasing the r600 on Agena on editor's day will have reviews that don't give the same shining outlook. Not to mention editor's day material will be under NDA until the reviews hit, so there will be almost no impact on the general consumer.

I agree but I think I'm coming around on the whole image thing. If AMD is able to demo a full suite of impressive product lines then that could restore some lost faith. At least they can start people talking about their imminent comeback in the CPU space. But like you said it will all mean nought for performance. R600 will be reviewed on Intel CPU's and there's nothing they can do about that.

Unknown Soldier

10-Apr-2007, 21:36

Which still has me wondering wth AMD held back R600. They had no reason to, except to launch the R6xx family and showing "Barcelona" that's gonna be 4 months or more late.

US

trinibwoy

10-Apr-2007, 21:51

Well the latest line we're being fed is that they would get better return on their marketing buck by doing it this way. Still sounds like a bunch of bull to me but I'm no CEO :)

Kaotik

10-Apr-2007, 22:02

Which still has me wondering wth AMD held back R600. They had no reason to, except to launch the R6xx family and showing "Barcelona" that's gonna be 4 months or more late.

US

Except the good PR image it gives when you show it on your hardware, I mean, how would it look if Ferrari would demonstrate their brand new revolutionary gearbox on a Mercedes rather than their own Ferrari?

icecold1983

10-Apr-2007, 22:06

if g80 refresh rly launches may 1st it will launch b4 r600. ati has fallen so low.

Voltron

10-Apr-2007, 22:49

if g80 refresh rly launches may 1st it will launch b4 r600. ati has fallen so low.

Could it be a repeat of the GF3 vs 8500 scenario where the Ti 500 served to keep the 8500 at bay until the GF4 arrived. That seems like a good bet to me, with a beefed up 65nm GF 8900 on the horizon - quite possibly this fall.

Shtal

10-Apr-2007, 23:00

Could it be a repeat of the GF3 vs 8500 scenario where the Ti 500 served to keep the 8500 at bay until the GF4 arrived. That seems like a good bet to me, with a beefed up 65nm GF 8900 on the horizon - quite possibly this fall.

I think those days are over :lol:
My believe that since ATI/AMD renamed R600 from X2800XTX to X2900XTX probably because it's not going to compete against 8800GTX but maybe will compete against 8900GTX.

Kaotik

10-Apr-2007, 23:46

I think those days are over :lol:
My believe that since ATI/AMD renamed R600 from X2800XTX to X2900XTX probably because it's not going to compete against 8800GTX but maybe will compete against 8900GTX.

The curious part is that they will "have to" name the refresh as X2950 now, which traditionally has been only for small changes, not a whole refresh

Razor1

10-Apr-2007, 23:56

True on one end, indeed, but it's no different from what Intel did with Core 2's, except for the video card involved this time

True, but in this case Penryn is there too, not to mention what ever nV has.

Strength to Wall Street and investors and banks, as nothing to do with an entire showcase of technology, it depends on competitiviness in each marketplace. If AMD demonstrates this kudos to them, if not, well things can't get much worse, or can they, many of the top fanacial advisors have already stated this entire year is bad for AMD, and there is doubt about next year as well since they will be sinking into further debt, even the 500 million in reduction of expenditure is not really much.

OEM's aren't going to care about a top to bottom launch, since the cpu and motherboards aren't going to be launched with the r600 family, so that has very little bearing.

This is a top to bottom "showcase" which I don't think its good, unless I'm underestimating the Barcelona core compared to Penryn, which I don't think I am, Penryn has a process advantage, which will allow for a clock rate boost if necessary and still stay in a competitive power envolope unlike presscott if needed.

nomedo

11-Apr-2007, 02:24

I think those days are over :lol:
My believe that since ATI/AMD renamed R600 from X2800XTX to X2900XTX probably because it's not going to compete against 8800GTX but maybe will compete against 8900GTX.

maybe they had to change that much of the R600 to make it keep up with the 8800GTX-oc cards that the name wasnt good enough.

or maybe they didnt want the press to put in a "L" in the old name and call it the x2L800-series (too late) :wink:

overclocked_enthusiasm

11-Apr-2007, 02:29

True, but in this case Penryn is there too, not to mention what ever nV has.

Strength to Wall Street and investors and banks, as nothing to do with an entire showcase of technology, it depends on competitiviness in each marketplace. If AMD demonstrates this kudos to them, if not, well things can't get much worse, or can they, many of the top fanacial advisors have already stated this entire year is bad for AMD, and there is doubt about next year as well since they will be sinking into further debt, even the 500 million in reduction of expenditure is not really much.

OEM's aren't going to care about a top to bottom launch, since the cpu and motherboards aren't going to be launched with the r600 family, so that has very little bearing.

This is a top to bottom "showcase" which I don't think its good, unless I'm underestimating the Barcelona core compared to Penryn, which I don't think I am, Penryn has a process advantage, which will allow for a clock rate boost if necessary and still stay in a competitive power envolope unlike presscott if needed.

Perhaps an early demo of Penryn by Intel to blunt the early showing of Barcelona? Gee...I wonder which viddy card they will use? Nvidia hardware is going to get the buzz with Penryn by default whenever it launches.

I see this as an "all in" by AMD at the poker table. The entire fate of AMD rests on the success of Barcelona vs Penryn or AMD will have to restructure completely. Their cash burn is anticipated at $1.5 billion this year which will leave them with zero in the bank and negative cashflow going forward. I highlight this to further my case of why AMD held R600 back to launch with Agena FX and RD 790. AMD must hit a homerun here on all three parts to sell their superior technology and regain mindshare or they may be forced to restructure.

Barcelona/Agena FX > R600

Sound_Card

11-Apr-2007, 03:13

Thanks to flippin waffles

http://forums.overclockers.co.uk/showthread.php?t=17712718

Hi there

Next week I shall be recieving one of the first R600 samples for testing.

We plan to put this card in our top Ultima specification of PC. We shall post pictures and tell you how good it is sometime next week.

However exact benchmark results cannot be revealed due to NDA but as you know I will give a subtle hint and post many pictures for you all.

Natoma

11-Apr-2007, 03:27

Perhaps an early demo of Penryn by Intel to blunt the early showing of Barcelona? Gee...I wonder which viddy card they will use? Nvidia hardware is going to get the buzz with Penryn by default whenever it launches.

I see this as an "all in" by AMD at the poker table. The entire fate of AMD rests on the success of Barcelona vs Penryn or AMD will have to restructure completely. Their cash burn is anticipated at $1.5 billion this year which will leave them with zero in the bank and negative cashflow going forward. I highlight this to further my case of why AMD held R600 back to launch with Agena FX and RD 790. AMD must hit a homerun here on all three parts to sell their superior technology and regain mindshare or they may be forced to restructure.

Barcelona/Agena FX > R600

Just so you know, AMD is already restructuring (http://money.cnn.com/2007/04/09/technology/amd.reut/index.htm) due to losses currently realized and anticipated. I think that makes this Barcelona/R600/RD790 launch all the more important.

Shtal

11-Apr-2007, 06:17

The curious part is that they will "have to" name the refresh as X2950 now, which traditionally has been only for small changes, not a whole refresh

I totally see your point!

Also It could be that R650 on 65nm tech be clocked very high and that's probably - be the only differentness from R600 "X2900XTX" is high clock speed, so it will carry X2950XTX name for Extreme High Clocks.

Geeforcer

11-Apr-2007, 06:33

Just so you know, AMD is already restructuring (http://money.cnn.com/2007/04/09/technology/amd.reut/index.htm) due to losses currently realized and anticipated. I think that makes this Barcelona/R600/RD790 launch all the more important.

The company in desperate need of money forgoes... errr... money by purposely delaying its product? That makes little sense to me.

Silent_Buddha

11-Apr-2007, 06:55

Or, company in need of cash figures to save a few million in advertising by launching/showing multiple products in the same event.

A few poster's above fail to realize that AMD is showing platforms/products to OEMs so that they can be used in systems for the upcoming back to school and holiday shopping seasons. Which is why I previously mentioned that this "launch" more so than any previous launch is focused on the OEM market. And to a lesser degree niche stream computing market. [edit] Which would mean that there's a good likelyhood that Athena and Athena FX will be available for OEMs to use in systems targeting one or both of those hot selling seasons.

If ALL they cared about was the enthusiast then they would have launched the R600 earlier, I'm sure. However, rather than looking at a short term band-aid to their cashflow, they are taking a longer more forward looking approach.

As to x2900xt, there's still the outside chance that R650 went so swimmingly well that it's what's being used for the high end. Thus the bump up in numbers as it is in effect the refresh of what should have launched last holiday season.

And R700 (if it's on time and has no snags) might presumably launch for the next holiday season. Thus rendering a need for a refresh part moot.

Granted I give all that less than a 10% chance of actually being the case, but you never know.

BTW - when does Q3 start for AMD? It's possible Athena FX might be closer than most are expecting.

And AMD has been in worse positions than what it finds itself in right now. Before Athon XP launched not only did they have negative cash flow they also had massive debt. And even then it took them a few years to high postive cash flow.

I actually think the company has a rosier outlook right now than they did when Intel pulled the plug on other companies being able to clone their chips.

After all I don't see AMD going back to the days of of having CPUs with somewhat competitive interger performance and less than half the floating point performance of Intel CPUs. :P

Regards,
SB
Regards,
SB

_xxx_

11-Apr-2007, 07:13

A few poster's above fail to realize that AMD is showing platforms/products to OEMs so that they can be used in systems for the upcoming back to school and holiday shopping seasons. Which is why I previously mentioned that this "launch" more so than any previous launch is focused on the OEM market. And to a lesser degree niche stream computing market. [edit] Which would mean that there's a good likelyhood that Athena and Athena FX will be available for OEMs to use in systems targeting one or both of those hot selling seasons.

Yeah, like without the joint presentation the OEM's would have a really hard time finding out about the existance or performance of those parts. How clever :)

icecold1983

11-Apr-2007, 07:25

ive yet to see one sensible explanation of how launching together possibly gives anyone more confidence in amd. all the people trying to back up this crazy idea remind me of the chewbacca defense from south park.

"IT DOES NOT MAKE SENSE"

http://www.youtube.com/watch?v=yU-tZy3NIS4

Rangers

11-Apr-2007, 07:28

Except the good PR image it gives when you show it on your hardware, I mean, how would it look if Ferrari would demonstrate their brand new revolutionary gearbox on a Mercedes rather than their own Ferrari?

Do you really think AMD cares about this?

All they have to do is make no recomendation what CPU to use, or even recommend only using AMD CPU's to test (because they're more energy efficient (which they are compared to Intel's current chips) or some made up reason), knowing full well intelligent reviewers will ignore it. Basically sweep it under a rug.

Otherwise, do you suggest AMD only ever release GFX cards when they happen to be the current CPU performance leader? How unrealistic is that? Even a scrooge would allow them to be behind 50% of the time, with no ill idea toward the company. After all, there ARE two main competitors here, ignoring that one is far better funded.

I mean hell, it's not like AMD didn't lead in performance for like the previous FIVE YEARS before the last six months, despite being a fraction Intel's size. Cut em a break.

Kaotik

11-Apr-2007, 08:14

Do you really think AMD cares about this?

All they have to do is make no recomendation what CPU to use, or even recommend only using AMD CPU's to test (because they're more energy efficient (which they are compared to Intel's current chips) or some made up reason), knowing full well intelligent reviewers will ignore it. Basically sweep it under a rug.

Otherwise, do you suggest AMD only ever release GFX cards when they happen to be the current CPU performance leader? How unrealistic is that? Even a scrooge would allow them to be behind 50% of the time, with no ill idea toward the company. After all, there ARE two main competitors here, ignoring that one is far better funded.

I mean hell, it's not like AMD didn't lead in performance for like the previous FIVE YEARS before the last six months, despite being a fraction Intel's size. Cut em a break.

Yes, I do believe they do care about it, on this specific launch - why? Because it's their first big launch after the merger, and they are trailing at the moment behind on both CPUs and GPUs

(and since when has AMD had anything as energy efficient as C2D's?)

Rangers

11-Apr-2007, 12:05

Yes, I do believe they do care about it, on this specific launch - why? Because it's their first big launch after the merger, and they are trailing at the moment behind on both CPUs and GPUs

(and since when has AMD had anything as energy efficient as C2D's?)

So what is Nvidia going to do if AMD regains the CPU lead at any point? Recommend reviewers review their products on AMD chips? Oh, the humiliation!

Also about the efficiency, I just thought I read somewhere (Dailytech/) that the 65nm X2's use less power than the C2D's..guess that doesn't speak to efficiency.

_xxx_

11-Apr-2007, 12:18

So what is Nvidia going to do if AMD regains the CPU lead at any point?

Absolutely nothing? Why should they even care, not being a CPU manufacturer?

erick

11-Apr-2007, 12:21

TheInq really dropped a bomb this time... either on their toes or on the world, time will tell:

http://www.theinquirer.net/default.aspx?article=38837

epicstruggle

11-Apr-2007, 12:29

TheInq really dropped a bomb this time... either on their toes or on the world, time will tell:

http://www.theinquirer.net/default.aspx?article=38837
Is there really anything new in there, outside of pricing?

erick

11-Apr-2007, 12:37

Rys

11-Apr-2007, 12:47

I'd say that a $400 price tag on a 65nm top of the line graphics card in 2007 is pretty outrageous, especially in the light of the nV's $999 8800 Ultra project.

And the other thing - when did we last see a highest-end GPU being produced in millions?
Why should the process affect the price too much? Production costs per chip are affected by die area, not transistor size (and generally the smaller the process the more expensive a wafer start is, especially on a foundry's brand new process). And are you 100% certain that $999 8800 Ultra exists?

erick

11-Apr-2007, 12:51

Hey, don't shoot the messenger. What I believe has nothing to do with this article. Just pointing out that the claims are quite incredible.

EDIT: Sorry, can't pipe down on this one:

Production costs per chip are affected by die area, not transistor size (and generally the smaller the process the more expensive a wafer start is, especially on a foundry's brand new process).

I'd say there's a pretty noticeable correlation between transistor size and die area. And the 65nm process has been around for a while... maybe not for ATI, but for AMD, sure. As for this $999 madness, I hope to God it isn't true.

Rangers

11-Apr-2007, 12:58

Absolutely nothing? Why should they even care, not being a CPU manufacturer?

It's their direct competitor?

I admit, it's not as bad as AMD using Intel chips, but it IS clearly a snicker type of situation.

Just imagine if Nvidia made CPU's too, and the situation was flipped, AMD did not make CPU's, so AMD needed it's GPU's to run on Nvidia CPU's. Think a few people would comment deragatorily on that? Yeah, they would.

_xxx_

11-Apr-2007, 13:07

Well the CPU part is definitely not a competitor for nV, though the company as a whole now is.

But since the reviewers will use the fastest CPU's available anyway, I don't see why this would be relevant? NV could care less if B3D, [H], Anand etc. test the cards with either CPU as long as thier chip shines.

nicolasb

11-Apr-2007, 13:23

Fudzilla latest....

65nm part coming:

http://www.fudzilla.com/index.php?option=com_content&task=view&id=472&Itemid=34

we can confirm that ATI is working on a smaller R600 marchitecture based chip codenamed R650. It is officially messed up with R600. The R600 chip is simply too hot and it has terrible heat dissipation and this problem has to be fixed. Dissipation of 230+ Watts is simply too much for any graphic chip and any standard.

We are seeing R520 got swapped with R580 situation all over again and this can definitely confirm that R600 won't have a good and great life. R600 will be replaced as soon as possible as soon as the R650 redesign is done. This is a very expensive but desperate move.

We don't know any schedules about R650 but we are sure that it comes later than Q2.(This is expanded upon significantly in the Inquirer link posted earlier).

12.4" and 9.5" cards use the same PCB:

http://www.fudzilla.com/index.php?option=com_content&task=view&id=471&Itemid=34

We had a chance to compare two cards, at least in measurements and we found out that 12.4 card PCB is almost the same like the 9.5 one. Both cards have circa 9.5 inch PCB with some different components.

This means that the cooler makes the size.

Picture of the power connectors on the 12.4" OEM card:

http://www.fudzilla.com/index.php?option=com_content&task=view&id=479&Itemid=34

One of the most interesting things will be the power consumption and we learned that 200+ watts number is not exaggerated. You can clearly see the two connectors, the first one having 2x4 pins and the second 2x3 pins, like the ones you can see on Geforce 8800 GTX or GTS cards.

Kaotik

11-Apr-2007, 13:53

There has been shots of those OEM boards with both 2x 6pin and 1x 8pin 1x 6pin configurations

Chris123234

11-Apr-2007, 14:10

I'd say that a $400 price tag on a 65nm top of the line graphics card in 2007 is pretty outrageous, especially in the light of the nV's $999 8800 Ultra project.

And the other thing - when did we last see a highest-end GPU being produced in millions?

A $400 Top end graphics card is an absolute impulse buy to me. Especially after being subjected to extremely high prices.

erick

11-Apr-2007, 14:57

A $400 Top end graphics card is an absolute impulse buy to me. Especially after being subjected to extremely high prices.

Exactly. I mean, they'll supposedly be giving away 8800 GTX performance on 8800 GTS prices. Who could resist?

Osamar

11-Apr-2007, 15:29

The shareholders? The graphic cards assemblers?

I think doctors should prescribe to depressed people anything they smoke at The Inq :shock:

trinibwoy

11-Apr-2007, 15:30

I admit, it's not as bad as AMD using Intel chips, but it IS clearly a snicker type of situation.

Well the difference is that AMD is primarily a CPU company and Nvidia is primarily a GPU company. So the "snicker" effect isn't as pronounced.

trinibwoy

11-Apr-2007, 15:38

As much as $400 cutting edge cards are great it seems doubtful to me. Even with a manufacturing cost advantage undercutting the current market by $200 will be hard to swallow. The chip isn't going to cost $200 less so who is going to absorb all that lost margin? And you would have to remove a performance point or two to squeeze enough product differentiation into $400.

And not to state the obvious but this only works for you if your flagship is as fast as the other guy's :)

erick

11-Apr-2007, 16:02

Well, for one they can really save money by using older GDDR3 memory. Remember the 512bit mem bus? It dictates that if ATI were to use the now relatively cheap 1,6GHz GDDR3 memory chips, they would still get 102 GB/s bandwidth. Not their top end material perhaps, but more than good enough for medium-to-high range cards.

The (probably dirt cheap by now) 1,4GHz GDDR3 would still net about 89,6 GB/s.

Bjorn

11-Apr-2007, 16:17

As much as $400 cutting edge cards are great it seems doubtful to me....

And not to state the obvious but this only works for you if your flagship is as fast as the other guy's :)

True. There would be absolutely no reason for ATI to lower the price that much unless the R600 has much lower performance then expected.

Jawed

11-Apr-2007, 16:25

The (probably dirt cheap by now) 1,4GHz GDDR3 would still net about 89,6 GB/s.
Definitely dirt cheap (PS3 and XB360 both consume a load of these, 256MB and 512MB respectively).

It's a very good point, though the cost of the board (all those layers), the cooler and the die mitigates it somewhat.

It would be interesting if a $350 X2900XL with 89.6GB/s of bandwidth appeared.

Jawed

INKster

11-Apr-2007, 17:15

R600 without heatsink:
http://www.fudzilla.com/images/stories/2007/April/graphics/r600_card.jpg

erick

11-Apr-2007, 17:22

Definitely dirt cheap (PS3 and XB360 both consume a load of these, 256MB and 512MB respectively).

It's a very good point, though the cost of the board (all those layers), the cooler and the die mitigates it somewhat.

It would be interesting if a $350 X2900XL with 89.6GB/s of bandwidth appeared.

Jawed

Even the cooler may be switched for less expensive versions as 65nm dies start rolling out. According to the article, AMD expects that going from 80nm to 65nm will reduce the power requirements (and thus heat dissipation) of the Rxxx series GPUs between 60-100W depending on the GPU. And there's a hell of a difference between 230W and 150W.

Jawed

11-Apr-2007, 17:32

Even the cooler may be switched for less expensive versions as 65nm dies start rolling out.
I was going to suggest this, but then remembered the rumoured dual-slot cooler on RV630XT, a 65nm GPU of ~160mm2 (>1/3 R600) with ~125W power consumption - actually the board will consume that much I expect, so perhaps the GPU only consumes about 100W or so.

The rumours relating to AMD's 65nm GPUs hardly indicate "low power" due to the process. But we don't know the performance levels of any of these GPUs (though the rumoured RV GPUs' bandwidths surely put a tight cap on performance).

But yeah, 65nm must result in cooler-running cores. But performance in the R650 refresh will have to be bumped up...

Jawed

Russell

11-Apr-2007, 17:54

I was going to suggest this, but then remembered the rumoured dual-slot cooler on RV630XT, a 65nm GPU of ~160mm2 (>1/3 R600) with ~125W power consumption - actually the board will consume that much I expect, so perhaps the GPU only consumes about 100W or so.

The rumours relating to AMD's 65nm GPUs hardly indicate "low power" due to the process. But we don't know the performance levels of any of these GPUs (though the rumoured RV GPUs' bandwidths surely put a tight cap on performance).

But yeah, 65nm must result in cooler-running cores. But performance in the R650 refresh will have to be bumped up...

Jawed

Unless Fudo's idea that they'll release the two together, with R650 as the highest-end part is true (hah!), then it won't have to be much faster.

But what are the odds that's true? *laugh*

The more I think about it though, I'm fairly certain these R650/R600 65nm rumors are abound for a very good reason. I think R650 just happens to be really damned close to being ready, due to the delays. Reminds me of R520/R580. I find that most rumors have a very good reason for being spread around, especially when they come up again and again. In this case, people are just confused over what's what and misread/misunderstand their source information.

EDIT:

Also, for those who keep saying that the R600 delay is costing AMD money...

From Le Inq: OEMs go nuts for DirectX 10 GPUs
AMD'S R6xx SERIES may be having a troubled birth, but the amount of these babies that are being sold numbers in the millions.

So they ARE selling chips, just not to YOU. I can understand the bitterness, but there you have it.

Geeforcer

11-Apr-2007, 18:54

TheInq really dropped a bomb this time... either on their toes or on the world, time will tell:

http://www.theinquirer.net/default.aspx?article=38837

I must say, this is the longest and most convoluted way of saying "We were wrong about R600 being a 65 nm part" imaginable.

_xxx_

11-Apr-2007, 19:00

Unless Fudo's idea that they'll release the two together, with R650 as the highest-end part is true (hah!), then it won't have to be much faster.

But what are the odds that's true? *laugh*

Call me nuts, but I could actually imagine that. It would make sense in some weird way.

Geeforcer

11-Apr-2007, 19:00

BTW, the image of R600 as troubled, semi-DOA part to be jettisoned at the earliest convenience CAN NOT inspire confidence in people who have been patiently waiting for it since November.

3dilettante

11-Apr-2007, 19:07

So they ARE selling chips, just not to YOU. I can understand the bitterness, but there you have it.

These must be presales or planned acquisitions (or the INQ is wrong). I don't know how they could ship millions of R6xx cores without there being a lot more info leaked.

Jawed

11-Apr-2007, 19:29

The more I think about it though, I'm fairly certain these R650/R600 65nm rumors are abound for a very good reason. I think R650 just happens to be really damned close to being ready, due to the delays. Reminds me of R520/R580.
I don't think it's unreasonable to expect 65nm R650 to be due for September/October - its original timing if R600 was supposed to debut in November 06.

So, ahem, May to September is 4 months. R650 has got be taping out now if that's its schedule.

Maybe, being a high-end part, its schedule is more November-ish. I dunno, I still find the timetables for these releases mystifying.

Maybe its original schedule was 6 months after R600, November->May? :lol: :lol: :lol: :lol: With RV6xx in February/March (but the RV6xx GPUs had bump or two on their road)?

Jawed

Jawed

11-Apr-2007, 19:30

These must be presales or planned acquisitions (or the INQ is wrong). I don't know how they could ship millions of R6xx cores without there being a lot more info leaked.
How much lead time before a launch does a value part enjoy? Does an inventory get built-up, or is on-going production in such high quantity that it's really only a matter of getting the production lines up to full speed?

How long does it take to get RV boards up to full production rate?

Jawed

3dilettante

11-Apr-2007, 19:44

I can't rightly say. I was just commenting on the distinction between selling a given item to an OEM versus its being sold to a consumer.
I'm under the impression that OEMs make big orders, and what one counts as a "sale" might be made prior to the product actually moving from a board manufacturer's inventory.

If millions were in OEM hands, there would be hundreds of test systems running, and right now not one of them apparently has had a curious employee play a game or take a picture of the insides.

trinibwoy

11-Apr-2007, 20:02

Order -> Purchase -> Receive.

Who knows what Fudo is talking about.

Razor1

11-Apr-2007, 20:23

http://www.nvnews.net/vbulletin/showthread.php?t=89589

some more pics

Creig

11-Apr-2007, 20:39

Samsung BC09 is GDDR4 rated at 2.2 Gbps

Jawed

11-Apr-2007, 20:42

I think this is the first absolute confirmation of double-sided 3-3-2 memory configuration - not surprising, but still comforting to see.

Jawed

DemoCoder

11-Apr-2007, 20:50

NocturnDragon

11-Apr-2007, 20:57

Geeforcer

11-Apr-2007, 20:58

I don't think AMD is this stupid and they are not in the business of Charity Giving, so either the price rumors are wrong, or the performance isn't that spectacular.

Looking at their strategy over the last year, I would not be so sure.

Silent_Buddha

11-Apr-2007, 20:59

IF OEMs actually are buying R600's in the millions.

And that's a pretty plausible "if" considering Apple appears committed to the chip which would indicate other OEMs would also be interested especially if R600 drivers really are as stable as AMD indicates. And OEM's more than anything want stable drivers to go along with hardware, thus their aversion to using G80 in high volume product lines.

Actually, the thought of "millions" of R600's going to OEM's does make sense. It solves a few problems.

1. Chips with higher than desirable heat output are relegated to a market that can afford to have rather large and exotic heatsinks capable of dissipating said heat.

2. It avoids the situation that ATI had with the launch of R420 where OEMs got large initial supplies but neither OEM's nor consumer's were able to get enough to satisfy demand due to splitting allocation between the two.

3. Large sales to OEMs is basically money in the bank. There is no gamble about whether enough will sell into the retail channel to regain lost expenditure. At this point the burden is on the OEM to sell machines with the parts, AMD already has the money.

4. Filling demand from OEMs first allows them to then stockpile parts for retail without having to worry as much about try to balance retail availability vs OEM availability. Related to point 2.

5. Possibly giving more time to manufacture 65nm R650 such that it may launch as the high end part rather than a "too hot" R600 part. I give this VERY low probability.

6. Of higher probability it would also allow them to weed out and stock up on cooler running parts for the retail market.

Being a pragmatic optimist I'll tend to take a company at its word until such time as they are proved to be wrong.

As such, this would also help to explain the further delay, when AMD states they could have launched R600 into the retail channel a month or more ago.

As to secrecy on the OEM's parts. Considering most are build to order machines, I don't imagine it takes a lot of lead time to install new graphics cards into existing machines. I wouldn't be surprised if, due to AMD's paranoid secrecy on R600 that they are required to store purchases in a warehouse until closer to official launch. With added price incentives to make sure they comply.

Also, if true, it would indicate a couple possible things.

1. R600 drivers are significantly more stable and problem free than G80 drivers. Thus, OEM's will continue to forgo offering G80s and just move right on to R600s.

2. R600 really IS priced significantly lower than the competition, thus again, motivating OEMs to bypass G80 in favor of R600.

Anyways, all that is just a running train of thought on a rumor that may or may not be true.

We'll all find out in a few weeks a little more of the truth behind things I'd suspect.

Regards,
SB

Silent_Buddha

11-Apr-2007, 21:10

To me, if you launch a card that is the absolute performance leader, or even on-par with the competition and sell it significantly cheaper, you made an idiotic business decision, because unless AMD has amazingly cheap manufacturing, and world-beating yields, then their costs are about the same. You sell your card very cheap when you don't have world beating performance and hope to take marketshare via price war, while suffering lower margins. But selling a world-beating performance card for less than the slower cards? I don't think AMD is this stupid and they are not in the business of Charity Giving, so either the price rumors are wrong, or the performance isn't that spectacular.

Except for one very important thing as it relates to AMD right now.

They are absolutely focused on maintaining and if possible increasing market share at this point.

One only has to look at the CPU side of things to see them sacrificing margin at the cost of trying to maintain market share.

Considering the momentum that Nvidia has and the marketshare they are gaining because of it, I'd imagine a 400 dollar price point would do a few things for AMD.

1. Regain some lost marketshare and stop Nvidia's momentum.

2. Hopefully (from AMD's point of view) make the few people that even know R600 is delayed forget about the delay or at least forgive them for it.

3. Generate good-will and massive word of mouth advertising from enthusiasts who would presumably jump on a chip with performance on par with a 550-600 dollar part from the competition.

4. Force Nvidia to respond in kind, thus eroding their margins.

Is it worth the loss of increased cash flow from a higher margin on the part? Noone will really know until it's done and people have had a few months to examine the impact of such a move.

It's a gamble pure and simple. It could either work or it could backfire. But in the business world you won't make significant strides unless you're willing to take the gambles that go with it.

Regards,
SB

Razor1

11-Apr-2007, 21:14

Deusp

11-Apr-2007, 21:33

neliz

11-Apr-2007, 21:33

SB the r600 costs around ~350 bucks to make (1 gig version) and 200-250ish for the 512 version, they would have no margins if they are to price it so low. That doesn't sound realistic if the r600 performs higher then g80. AMD doesn't have much to go by right now and not cut into more profits.

Ahem.. wasn't everyone talking about mindshare?
Their no-margin low-volume unit is taking care of the mouth-to-mouth advertisement and generating a positive mindset (look nV's $1000 part is only marginally faster than this $400 board) and reflect positive radiance on the high-margin high-volume products.

It looks like a risky investment from AMD, but if you don't have money, why don't you generate marketing from sales?

BTW, WHERE are the benchmarks eh?

Deusp

11-Apr-2007, 21:40

Except for one very important thing as it relates to AMD right now.

They are absolutely focused on maintaining and if possible increasing market share at this point.

I can't believe a company that's losing as much money as AMD is right now is thinking that.

One only has to look at the CPU side of things to see them sacrificing margin at the cost of trying to maintain market share.

That's because Intel has the better product on the market currently. If Intel cuts prices, AMD must follow suit or be blown away.

Considering the momentum that Nvidia has and the marketshare they are gaining because of it, I'd imagine a 400 dollar price point would do a few things for AMD.

1. Regain some lost marketshare and stop Nvidia's momentum.

2. Hopefully (from AMD's point of view) make the few people that even know R600 is delayed forget about the delay or at least forgive them for it.

3. Generate good-will and massive word of mouth advertising from enthusiasts who would presumably jump on a chip with performance on par with a 550-600 dollar part from the competition.

4. Force Nvidia to respond in kind, thus eroding their margins.

Is it worth the loss of increased cash flow from a higher margin on the part? Noone will really know until it's done and people have had a few months to examine the impact of such a move.

It's a gamble pure and simple. It could either work or it could backfire. But in the business world you won't make significant strides unless you're willing to take the gambles that go with it.

Regards,
SB

Like I said, it will only force themselves into a price war with a company in a better financial state. It's a gamble that's almost certainly going to backfire and quite likely to put them out of business before they can see to its completion.

Razor1

11-Apr-2007, 21:43

Ahem.. wasn't everyone talking about mindshare?
Their no-margin low-volume unit is taking care of the mouth-to-mouth advertisement and generating a positive mindset (look nV's $1000 part is only marginally faster than this $400 board) and reflect positive radiance on the high-margin high-volume products.

It looks like a risky investment from AMD, but if you don't have money, why don't you generate marketing from sales?

BTW, WHERE are the benchmarks eh?

what is the lowest margin high end card we have seen? The x1900xtx comes to mind, and they had a ok margins before the gf8's came out. How much marketshare is price going to garner if it forces people to buy new power supplies? Doesn't seem like a cost beneficial thing unless they start bundling PSU's with these things.

Is the x1900 series increasing mindshare, yes in the lower upper range, but only a few % points. 3-4%. Thats not enough to make much of a difference if they are going to drop their pants to lose 20-25% margins. At the top end its not even 3-4% its more like 2%-3%.

neliz

11-Apr-2007, 21:50

What if the low margin card, the 512MB GDDR3 version was on level with the GTX?

If you can run a GTX, you can run a X2900XT, there is like 20Watt theoretical difference between the two parts.

And the part that does sell for, $549 or whatever gives you the performance crown. you only need to pay the difference between the two versions for a potent PSU.

Razor1

11-Apr-2007, 22:10

thats very possible, I would think thats the route AMD is going to take.

erick

11-Apr-2007, 22:46

I think people are overlooking the fact that the high-end card does not necessarily need to make money. You can settle for just covering the costs involved, IF it helps you regain mindshare+marketshare and sell loads of cheaper, lower-performing cards.

Or better yet, if you already have sweet deals for those low-end parts (according to that TheInq article the RV610 is quite a hit with OEMs like Dell), then you won't risk anything more than temporarily losing some of your profit or, effectively, trading it in for more marketshare.

11-Apr-2007, 23:16

http://we.pcinlife.com/thread-747857-1-1.html

Pictures are already removed, but one supposedly showed a R600 scoring 11K at 1600x1200 in 3DM06.

Edit: fashionably late. same link as the one on NV News. :P

DemoCoder

11-Apr-2007, 23:19

Apple isn't in the business of buying "millions" of GPUs for Mac Pros, nor does Apple manufacture cards either. The Mac Pro isn't selling well, definately not in the millions.

The "R600 2900XT(X) is a loss leader to sell low/midrange" doesn't seem to be to be all that smart. Historically, IHV's have used the high-end cards to promote their leadership image to try and burnish their image with OEMs, but these cards typically have high margins and were sold at a premium, which paradoxically sometimes helps to increase one's image (Quadro anyone? 50 cent t-shirts sold for $50, etc)

The R600 isn't a console, where they expect to get licensing revenue from software by selling hardware at a loss. In fact, a user who buys an R600 "at a loss" isn't likely to buy a low/mid range ATI card too ala 'attachment rate' we are so fond of, so the high end card acts to cannibalize the sale.

At best, one could somehow hope that oodles of high end gamers with R600's would convince Joe-Blow to buy low end ATI cards. But if that's the strategy, why not just throw a LAN Party event and hand out a few thousand cards, or sell them at a loss *at the LAN event* for the biggest media "oomph"

I don't buy it. AMD desparately needs to increase its margins.

Unknown Soldier

11-Apr-2007, 23:20

Except the good PR image it gives when you show it on your hardware, I mean, how would it look if Ferrari would demonstrate their brand new revolutionary gearbox on a Mercedes rather than their own Ferrari?

Sorry, we aren't talking cars here. These are GPU's which will be installed on Intel Hardware whether AMD like it or not. Holding the GPU's(R600) back so that they could be displayed on AMD's Barcelona which won't even see the day of light(for customers) until another 3-4 months down the line and will most probably be displayed on Intel's new QX6800 the day it does get reviewed by 90-95% of the sites, just goes to show that either AMD are arrogant or that someone made a big mistake in March in regards to both the GPU and CPU. That said, if Barcelona still comes out using higher watts that the Intel Quad but are only able to be as fast as the Quad's, I expect most people will stick to the Intel's.

And for your info, Ferrari allow Spyker to use their engine and Williams has allowed Toyota to use their seemless gearbox.

US

_xxx_

11-Apr-2007, 23:20

I think people are overlooking the fact that the high-end card does not necessarily need to make money. You can settle for just covering the costs involved, IF it helps you regain mindshare+marketshare and sell loads of cheaper, lower-performing cards.

This is really becoming the new myth. Where do you people get that sort of stuff? You may launch a part with some loss, but that will be limited according to thorogh calculations and planning. It's not like "hey, let's put millions into this part and then sell it with a loss to make the masses like us" or such ;)

Frank

11-Apr-2007, 23:29

How fast is the RSX compared to the 8800? Because the Xenos is doing a bit better than the RSX it seems. And it's pretty certain that the R600 will be a bit bigger Xenos. That only leaves the question if the RSX is closely related to the 8800. Which seems to be the case.

bigtabs

11-Apr-2007, 23:29

This is really becoming the new myth. Where do you people get that sort of stuff? You may launch a part with some loss, but that will be limited according to thorogh calculations and planning. It's not like "hey, let's put millions into this part and then sell it with a loss to make the masses like us" or such ;)

Funny how you quoted him, yet completely misrepresented what he said. :wink:

bigtabs

11-Apr-2007, 23:31

Twinkie

11-Apr-2007, 23:34

I think the RSX is more like a 7600GT no?

Its more of a 90nm 7800GTX clocked at 550Mhz.

edit - with abit of modification.

Frank

11-Apr-2007, 23:34

I think the RSX is more like a 7600GT no?
That's the big secret. It is assumed, yes. And that's about what we know about it.

INKster

11-Apr-2007, 23:43

That's the big secret. It is assumed, yes. And that's about what we know about it.

It is a bit big for a 7600 GT, no ?

Since it (RSX) has close to 300 Million transistors and the 7600 GT (G73) only has ~170 (even taking into account the different metric of transistor count between TSMC's 90nm and Sony's own 90nm processes), you must wonder what are those extra units for, especially when we already know that some 20M (the Purevideo processor) are not present at all in Sony/Nvidia's RSX.

People assume RSX is a close relative of the 7600 GT simply because of an abstract number such as the width of its memory bus being the same (128bit).
There's more to it.
The 128bit 7600 GT has no problem beating a 6800 Ultra with its 256bit bus, for instance.

Unknown Soldier

11-Apr-2007, 23:46

http://www.nvnews.net/vbulletin/showthread.php?t=89589

some more pics

I don't see any audio connector(or HDMI for that matter) in those pics. Of course that is the OEM version which should be 80nm. Don't know how packed the retail version is gonna look like with that audio(or HDMI connector). So nothing really new to what has already been shown 50000 times on the net since January(maybe except the memory ratings and configuration).

US

Kaotik

11-Apr-2007, 23:51

It's only the reference design, I'm guessing it will be up to the AIB's to decide wether to put 2x DVI or 1x DVI 1x HDMI there, or maybe even 2x HDMI?
I mean, there never was a reference model for the X1600's and whatever with HDMI, but they actually did ship them with 1xDVI + 1x HDMI configurations too

Arnold Beckenbauer

12-Apr-2007, 00:05

I don't see any audio connector(or HDMI for that matter) in those pics. Of course that is the OEM version which should be 80nm. Don't know how packed the retail version is gonna look like with that audio(or HDMI connector). So nothing really new to what has already been shown 50000 times on the net since January(maybe except the memory ratings and configuration).

US

What? You haven't read Fudo's sentional news?
http://www.fudzilla.com/index.php?option=com_content&task=view&id=485&Itemid=1
I have no idea, how this could work. Some kind DVI-IA "integrated audio"?

Unknown Soldier

12-Apr-2007, 00:18

Oh wow .. ok!

US

Russell

12-Apr-2007, 00:29

Call me nuts, but I could actually imagine that. It would make sense in some weird way.

AMD doing something unexpected or Fuad being right? :)

These must be presales or planned acquisitions (or the INQ is wrong). I don't know how they could ship millions of R6xx cores without there being a lot more info leaked.

If they have millions of cards ordered, surely some money has already changed hands. That's all I was trying to say.

I don't think it's unreasonable to expect 65nm R650 to be due for September/October - its original timing if R600 was supposed to debut in November 06.
Jawed

I agree that a early autumn release is most likely, but all the rumors of a 65nm "R600" part taping out are too fresh in my mind. I'm talking about the ones prior to the recent "OMG 65NM!!!" phase, but the rumors back in January of a 1GHz 65nm chip. I wouldn't be surprised to see this chip tailing R600 but an uncomfortable margin, though they could delay it to make that a bit happier.

bigtabs

12-Apr-2007, 00:48

Rather than delay, why not just bump up the price?

INKster

12-Apr-2007, 01:03

What? You haven't read Fudo's sentional news?
http://www.fudzilla.com/index.php?option=com_content&task=view&id=485&Itemid=1
I have no idea, how this could work. Some kind DVI-IA "integrated audio"?

To my knowledge, DVI-I can't carry digital audio without breaking the specs, and then it would not be called DVI anymore... :wink:

Sobek

12-Apr-2007, 01:18

Originally Posted by Razor1
http://www.nvnews.net/vbulletin/showthread.php?t=89589

some more pics

Is it wrong that this arouses me?

Now all we need are the original images in their highresolution format. In this day and age, no one takes such photos at anything less than 3000x2000 :razz:

Razor1

12-Apr-2007, 01:29

Is it wrong that this arouses me?

Now all we need are the original images in their highresolution format. In this day and age, no one takes such photos at anything less than 3000x2000 :razz:

size means everything some people, lucky my ex girlfriends didn't feel this way :oops:

jamis

12-Apr-2007, 01:34

size means everything some people, lucky my ex girlfriends didn't feel this way :oops:
It's just what they say, afterall they are all ex... :wink:

Cuthalu

12-Apr-2007, 02:52

So, after all, it's as fast as G80 @ 1280*1024 at 3dmark? Not very bad, but someone should bench real games.

trinibwoy

12-Apr-2007, 03:00

They're also completely forgetting about game theory, namely that nVidia won't sit still against this.

I see somebody is familiar with basic micro-economics :)

Of course, you're right and the only way AMD can avoid this trap is to have phenomenally lower costs which is highly unlikely. But I believe the $400 R600 rumour about as much as I believe the $1000 G80 one.

That sure is an impressive looking card though, looks like the light at the end of the tunnel is fast approaching.

Razor1

12-Apr-2007, 03:45

It's just what they say, afterall they are all ex... :wink:

heh ya got the secondary joke :grin:

well back on topic, so hopefully some benchmarks will be leaked soon :)

Cuthalu

12-Apr-2007, 03:53

well back on topic, so hopefully some benchmarks will be leaked soon :)

In case you missed it, CJ already spotted something:
http://we.pcinlife.com/thread-747857-1-1.html

Pictures are already removed, but one supposedly showed a R600 scoring 11K at 1600x1200 in 3DM06.

Razor1

12-Apr-2007, 03:56

In case you missed it, CJ already spotted something:

they didn't have anything about performance, I translated that page and they only mention physical aspects of the card. Each picture has a discripition next to it.

Sobek

12-Apr-2007, 04:02

they didn't have anything about performance, I translated that page and they only mention physical aspects of the card. Each picture has a discripition next to it.

There may have been something on one of the images that OP never intended (or fully intended) to be seen that wasn't included in the description.

Cuthalu

12-Apr-2007, 04:06

At least there's some speak about it on the comments: "这卡3DMARK06 1600*1200跑11000跟好玩一样 ", for example.

Razor1

12-Apr-2007, 04:07

number of missing photos and number photos at nvnews line up too 15 to 15

Well if it is 11k, thats around 20% higher then a g80 at that res so its possible.

Pete

12-Apr-2007, 04:39

That's the big secret. It is assumed, yes. And that's about what we know about it.
As Twinkie said and INKster implied, RSX is closer to a 7900 than a 7600. It has the shader units of a 7900 (24) and the bus width of a 7600 (128bit, ignoring its conxn to Cell's XDR RAM).

That said, I have no idea how relating Xenos to RSX and the 8800 is relevant to this thread, but I guess it's about the only avenue these R600 threads haven't explored....

Anyway, speaking of those PCInlife/NVN pics, can we more accurately guesstimate R600's power draw by analyzing the power circuitry near the R600 board's, erm, stern? Xbit says (http://www.xbitlabs.com/articles/video/display/ati-x1950xtx_5.html) I'm talking about the MOSFETs and induction coils (here's info (http://www.electronicstalk.com/news/pls/pls142.html) about the Pulse PA131xNL series of inductors, of which the R600 has a 1312 and 1314). I just remember the various R5x0 series high ends sporting varying #s of them.

silent_guy

12-Apr-2007, 07:14

I think this is the first absolute confirmation of double-sided 3-3-2 memory configuration - not surprising, but still comforting to see.

And at least we now have final confirmation that there is at least one version that's using GDDR4. They're using 2.2GHz types. (Samsung K4U52324QE-BC09 (http://www.samsung.com/Products/Semiconductor/GraphicsMemory/GDDR4SDRAM/512Mbit/K4U52324QE/K4U52324QE.htm))

_xxx_

12-Apr-2007, 07:30

Funny how you quoted him, yet completely misrepresented what he said. :wink:

What did I misrepresent? He said that high-end cards don't need to make money, I say that's bollocks.

_xxx_

12-Apr-2007, 07:32

AMD doing something unexpected or Fuad being right? :)

Both :)

I actually meant the "mixed" release of 65 and 80 nm parts.

Rangers

12-Apr-2007, 07:41

I can't believe a company that's losing as much money as AMD is right now is thinking that.

That's because Intel has the better product on the market currently. If Intel cuts prices, AMD must follow suit or be blown away.

Like I said, it will only force themselves into a price war with a company in a better financial state. It's a gamble that's almost certainly going to backfire and quite likely to put them out of business before they can see to its completion.

You just described a recipe such that, the smaller (of two hypothetical competing) companies must ALWAYS either be faster, or go out of business.

Clearly that's not the case, as AMD survived for a long time without doing either.

It's possible for example, for the smaller company to survive comfortably long term on lower margins than the larger company is willing to match (never mind the possibility of better operating efficiency as a corporation, which basically, is what decides success or failure in multi billion dollar business on a daily basis). As well as of course, tons of other reasons, including lack of capacity to satiate all demand.

Anyways, 20% faster, if the leaked bench is true, is a lot and I think personally, would be "enough" to make R600 a success. It's right on the borderline, but I think it crosses it.

erick

12-Apr-2007, 07:57

What did I misrepresent? He said that high-end cards don't need to make money, I say that's bollocks.

No, your comment read like I was talking about how its okay for ATI to lose money on R600, which I absolutely did not. When I said that its okay to settle for covering all costs involved, this is exactly what I meant. Breaking even. Not making a profit, but not making a loss either.

Besides, if the yields get better (and they certainly will), the profit per chip is bound to increase as the cost of working R600/R650 chips goes down. Then its up to AMD to decide if they wish to make the pricing even more competitive or start making some profits.

_xxx_

12-Apr-2007, 07:59

No, your comment read like I was talking about how its okay for ATI to lose money on R600, which I absolutely did not. When I said that its okay to settle for covering all costs involved, this is exactly what I meant. Breaking even. Not making a profit, but not making a loss either.

Well then it was a slight misunderstanding, I quoted that since there were many people around parroting that the high end stuff always loses money in order to sell the low-end, regardless of the company.

nutball

12-Apr-2007, 08:04

I think people are overlooking the fact that the high-end card does not necessarily need to make money. You can settle for just covering the costs involved, IF it helps you regain mindshare+marketshare and sell loads of cheaper, lower-performing cards.

If you reduce the price of the high-end card from $600 to $400 (for example), what happens to the price of the card which used to cost $350? $300? $250? $200? And so on. You'd be compressing the entire price range and/or reducing the selling prices of the cards where you really do make most of your money (the mid- and low-end). Hell, you may even end up tempting more people to buy your top-end, loss-making card (losing you more money in the process).

erick

12-Apr-2007, 08:23

If you reduce the price of the high-end card from $600 to $400 (for example), what happens to the price of the card which used to cost $350? $300? $250? $200? And so on. You'd be compressing the entire price range and/or reducing the selling prices of the cards where you really do make most of your money (the mid- and low-end). Hell, you may even end up tempting more people to buy your top-end, loss-making card (losing you more money in the process).

I feel that the current product ranges are overpopulated as it is. Having 8+ cards on your product range is clearly the result of over-the-top pricing for your high-end cards, since you need something to fill the gaps between $75 and $549.

If your top-of-the-range card costs $400 you need much fewer price range fillers, thus making for a more cohesive and compact product line. If you don't want your high-end stuff to cannibalise your lower-performing parts, than make sure there's enough of a price difference between the them. Again, that's not a problem if you've got 4 to 5 cards, but it would be a problem if you had 8 cards for example.

As a personal note - I actually dare not hope for a $400 R600, but I think a $450 part is quite realistic.

_xxx_

12-Apr-2007, 08:30

I'm still leaning towards $599.

nutball

12-Apr-2007, 08:36

Having 8+ cards on your product range is clearly the result of over-the-top pricing for your high-end cards, since you need something to fill the gaps between $75 and $549.

I'm not sure I agree. I think it's the other way round, I think having a $549 priced top-end card is the result of wanting to have 8+ cards in the range.

ANova

12-Apr-2007, 08:47

Ah the good old days of the Voodoos, Geforce 3s and 4s and 9x00s where there were only 2-4 models.

Kaotik

12-Apr-2007, 09:12

Ah the good old days of the Voodoos, Geforce 3s and 4s and 9x00s where there were only 2-4 models.

I'd say nVidia swayed off those "good old days" with GF4's, with the 3 models + 2 MX's (excluding the later released versions, there was in total 5 or 6 GF4Ti's and few more MX's too I think) which had nothing to do with their "bigger brothers" in terms of technology, MXes being GF2's

ANova

12-Apr-2007, 09:38

There were only three models of Geforce 4 Ti's, 4200, 4400 and 4600; later they expanded it to 2 or 3 more models with minor variations in clock/memory or to support AGP 8x. Like you said the MXs were essentially just rebranded GF2s, the performance gap between them and the GF4s were quite large.

Russell

12-Apr-2007, 10:01

Fuad
It looks like AMD will have to skip one more deadline. It is likely that it won't launch the RV630 and RV610 alongside the highly anticipated R600. Radeon X2900 series will officially launch in May and the NDA presentation will take place in Tunisia on the 23rd and 24th of April.

http://www.fudzilla.com/index.php?option=com_content&task=view&id=499&Itemid=1

I noticed that Fuad has a trend lately. Instead of saying ridiculous things that he will later get mocked for, his predictions are based on what is likely to happen. In this case, I wouldn't be surprised if he were spot on. I think he mentioned this one a couple weeks ago too, but is "Fuad-confirming" it now.

Yep. If this is true it's pretty shitty.

Geeforcer

12-Apr-2007, 10:25

Fuad

http://www.fudzilla.com/index.php?option=com_content&task=view&id=499&Itemid=1

I noticed that Fuad has a trend lately. Instead of saying ridiculous things that he will later get mocked for, his predictions are based on what is likely to happen. In this case, I wouldn't be surprised if he were spot on. I think he mentioned this one a couple weeks ago too, but is "Fuad-confirming" it now.

Yep. If this is true it's pretty shitty.

Of course, if this is true (mammoth-sized "if", source considered), then AMD will without a doubt delay R600 as well, right? I mean, they have been tripping over themselves telling people how R600 has been ready to go for months if not years (decades?) and they only reason it has not seen the light of day yet is due to how orgasmotastic "family launch" is.

nicolasb

12-Apr-2007, 11:00

I don't think it's unreasonable to expect 65nm R650 to be due for September/OctoberAren't we expecting R700 to be out before the end of this year? I think the original plan was to launch R600 last November/December, with R700 to follow a year later. That would make the likely original plan for an R650 release Summer rather than Autumn/Fall - presumably hot on the heels of the smaller 65nm chips. If that's correct and R650 is still on schedule then it could quite easily ship in June or July rather than October.

As far as the "high end cards will cost $400" rumour, even AMD isn't stupid enough to sell their high-end chips at a loss. There are therefore only two possibilities:

1) The rumour is false.
2) There will be a huge cost advantage over rival products. (Not sure where this might come from. 65nm vs 80nm, presumably.)

If there is actually a major cost advantage then selling cheap would make sense. If absolutely nobody buys the rival Nvidia product, then sales double at a stroke. And if the price of a high-end card drops from $600 to $400, the number of potential buyers of the high-end card is dramatically increased. It would be quite reasonable to sell the chips at a reduced margin if, by doing so, the cost of the eventual cards can fall by that much. But the suggestion that they could be sold at a loss is preposterous.

vertex_shader

12-Apr-2007, 11:26

Of course, if this is true (mammoth-sized "if", source considered), then AMD will without a doubt delay R600 as well, right? I mean, they have been tripping over themselves telling people how R600 has been ready to go for months if not years (decades?) and they only reason it has not seen the light of day yet is due to how orgasmotastic "family launch" is.

OEM has higher priority than retail, OEM's scared about NV vista drivers, so AMD has the chance now for the big "fishes".
OEM release is a launch too, looks like its coming in time, but its will be a dissapointment for the users who are waiting for retail amd dx10 mainstream cards.

Razor1

12-Apr-2007, 11:58

Definitely dirt cheap (PS3 and XB360 both consume a load of these, 256MB and 512MB respectively).

Jawed

Hmm funny price of gddr3 has been pretty much stable for the past few months to a year, you might want to look that info up again Jawed.

Arun

12-Apr-2007, 12:09

Hmm funny price of gddr3 has been pretty much stable for the past few months to a year, you might want to look that info up again Jawed.Source? Unless you asked Samsung directly, which can definitely work if they're in a good mood, I fail to see how you can say that.

Razor1

12-Apr-2007, 12:14

supply and demand, overall demand has increased with next gen products for GDDR3 memory specially even the lower mid range cards and all the way up are going to be using it plus the two consoles. Unless I'm missing something has Samsung increased production in the past year for GDDR3, maybe openned up a new fab or two we don't know about? Remember the court case about DDR ram and stabilization of prices that was only 6 months ago, and actually prices went up a little after that.

Unknown Soldier

12-Apr-2007, 12:32

OEM has higher priority than retail, OEM's scared about NV vista drivers, so AMD has the chance now for the big "fishes".
OEM release is a launch too, looks like its coming in time, but its will be a dissapointment for the users who are waiting for retail amd dx10 mainstream cards.

Ye, i'm beginning to think the real reason why AMD never launched the R600 in March was because Dell(and other OEM's) bought all the chips and placed orders for the first batch of RV610/630's and thus AMD has now supplied them or is still delivering on the 'Million plus chips' (http://www.fudzilla.com/index.php?option=com_content&task=view&id=499&Itemid=34) that is supposed to be delivered and not the original rumours that floated about saying that the chip/card had problems.

That would make AMD delay the release of the chips to retail and get them working on a power(retail) friendly solution(R650) in the meantime until the actual time as they have stock for retail purposes and a proper launch hence the new delay.

Hmm.. now i'm thinking that AMD might actually have done something right in order to keep the company above the water. For us consumers though it's a bitter pill but I guess that's where the money lies.

Now i'm beginning to think that AMD might actually release the 65nm R650 for retail consumers at release date instead of the 80nm R600 which they'll sell to the OEM's before hand.

US

trinibwoy

12-Apr-2007, 12:55

http://www.fudzilla.com/index.php?option=com_content&task=view&id=506&Itemid=1

ATI won't be using X in its Rx6xx brand names. The new Radeon generation will be branded Radeon HD 2000 series.

The R600XT or Radeon HD 2900HD is actually a 512 MB GDDR 3 card and the XTX will come later.

All the Radeon Rx6xx chips will be known as ATI Radeon HD 2000 Series.

I like it !! I hated that extra X :)

IbaneZ

12-Apr-2007, 12:59

Hi there

It arrived at 10:30am this morning, had to sign some NDA forms meaning I cannot disclose to much info.

Initial thoughts are it looks good, its very silent. Having issues at the moment with the alpha/beta drivers and no doubt due to the fact the OS had a NVIDIA card previously installed.

But on noise its quiet, pretty much silent. It runs hotter than a GTX but does not burn to the touch like previous ATI cards.

I will report back later in regard to performance once I've got past installation gremlins.

Gibbo got "his" R600.

http://forums.overclockers.co.uk/showpost.php?p=9109924&postcount=490

vertex_shader

12-Apr-2007, 13:10

http://www.fudzilla.com/index.php?option=com_content&task=view&id=506&Itemid=1

I like it !! I hated that extra X :)

Very good name choice :smile:

vertex_shader

12-Apr-2007, 13:11

Gibbo got "his" R600.

http://forums.overclockers.co.uk/showpost.php?p=9109924&postcount=490

It has a 8-pin and 6-pin PCI-E connectors but you can plug a 6-pin connector into the 8-pin without issues so you won't need to upgrade your PSU's to the new types.

So the power consuption less than 225watt.

Kaotik

12-Apr-2007, 13:29

http://www.fudzilla.com/index.php?option=com_content&task=view&id=506&Itemid=1

I like it !! I hated that extra X :)

You can be damn 100% sure it won't be called "Radeon HD 2000 HD" :lol:
And hasn't AMD itself been already referring to it as X2900? :???:

Sound_Card

12-Apr-2007, 13:30

Power consumption has to be less than 225w if you can use two 6-pin connectors.

Sound_Card

12-Apr-2007, 13:31

You can be damn 100% sure it won't be called "Radeon HD 2000 HD" :lol:
And hasn't AMD itself been already referring to it as X2900? :???:

I actully like the sound of x2900hd

Razor1

12-Apr-2007, 13:40

Hi there

Well got a few things running, though I am not confident at all with the drivers as things do see glitchy, might possibly need a fresh OS installation.

Am impressed though by the fact whilst underload it still remains nice and quiet, the fan control seems to work quite well, certainly a lot better than the old X1950's which seem to be just either on or off with the volume so the improved fan control is an nice improvement as it means now whilst gaming it still remains quiet.

Performance can't really say but at the moment its a mixed bag, some things its very impressive and others am unsure about but the image quality does seem some what better than the GTX with certain settings which could account for the lower FPS on some test, just seems more sharper and more details thrown into the image.

All I can say is wait for official reviews to get the real info on benchmarks and how it performs. For the moment I can not come to a conclusion on the performance until I've had more time with the card.

I feel it will be around GTX performance but has more features, better Vista performance and cost quite a lot less than a GTX. http://forums.overclockers.co.uk/images/smilies/smile.gif

hmm looks like Gibbo ain't goin give much.

Evildeus

12-Apr-2007, 13:44

cost quite a lot less than a GTX. hmmm

Jawed

12-Apr-2007, 13:44

Aren't we expecting R700 to be out before the end of this year? I think the original plan was to launch R600 last November/December, with R700 to follow a year later. That would make the likely original plan for an R650 release Summer rather than Autumn/Fall - presumably hot on the heels of the smaller 65nm chips. If that's correct and R650 is still on schedule then it could quite easily ship in June or July rather than October.
I've always thought that rumour about R700 was rot. A year between major architectural changes doesn't seem reasonable. Unless R700 is much simpler. I admit the rumour about what R700 is (multi-core) seems a bit far-fetched, so it could be simpler.

It's worth pausing on the fact that Orton actually mentioned R700 during that RD690 press event video.

Jawed

pakotlar

12-Apr-2007, 13:44

This doesn't sound too great. While I will conceede that newer, more stable, and better performing drivers may be available on launch, I was hoping that the initial impression Gibbo got would be a bit better. Considering how late it is, I was hoping drivers would be in a less "glitchy" state. I guess corporate restructuring could have affected development. I may still consider it, but my x1900xt will serve me a while longer i think.

Though this may be a similar case to x1800xt, where driver maturity REALLY affected performance. I bet scheduling is a whole lot more complex on this card than in the x1800 and x1900 cases. If we see similar boosts with this card that we saw through the duration of the x1800/x1900 product cycle (up till now), performance may still have headroom. Though I wonder what "similar" performance entails. If it is similar under cpu or fillrate limited scenarios, I won't be dissapointed, as long as the "impressive" scenarios are cases where such bottlenecks go away. I mean I want to see superior math performance, and handling of certain effects due to high bandwith, etc. Utilize your strengths ATI!

edit: As an afterthought, ATI sure has some tough competition. Nvidia has been so on the ball since NV40, that each successive generation improves not only in performance but design principles. I really feel like nvidia delivered the total package with the 8800GTX. Its the first part that I've been really compelled to switch for since NV25. IQ is very important to me, and for a while now, ATI had a major lead. Not the case anymore.

Jawed

12-Apr-2007, 13:57

So the power consuption less than 225watt.
But if he's only got XT, not XTX then we don't know so much...

Jawed

Razor1

12-Apr-2007, 14:03

Bouncing Zabaglione Bros.

12-Apr-2007, 14:05

Now i'm beginning to think that AMD might actually release the 65nm R650 for retail consumers at release date instead of the 80nm R600 which they'll sell to the OEM's before hand.

US

Welcome to the dark side.

Razor1

12-Apr-2007, 14:06

I would take that with a grain of salt for sure. As for the fan noise, if he is right, I will be extremely impressed that the card can remain quiet under full load considering it has to dissipate 250w of heat. Moving that kind of heat is going to take a lot of airflow IMO. As for his thoughts on performance, I am hearing that it would be hard to differentiate the R600 and GTX in a Pepsi Challenge.

Kyle just gave some specifics on performance some what, still a bit weary about his power numbers though.

nicolasb

12-Apr-2007, 14:08

The extra ram itself will eat up a good deal doesn't it? So even if its a little higher then 225 the core might not be that bad. But as you said if its the xt then its not good, but I don't know if Gibbo would make a mistake of not knowing what he has, wait did he mention if it had 1 gig of ram?If the card can accept two 6-pin PCIe connectors then it's physically impossible for the chip, memory, cooler and every other thing on the card combined, to draw more than 225W even at the very hightest instantaneous peak of power usage. In fact, there has to be a little margin for error, so the peak usage cannot ever actually hit 225W.

At the AMD press conference a figure of approximately 200W was mentioned - I see no reason to doubt this.

vertex_shader

12-Apr-2007, 14:10

hmm looks like Gibbo ain't goin give much.

What part you write this conclusion?

They write Better IQ, more features, less price, better vista performance, fan quiet in load, for me looks like the xt will be a hot seller (if true what Gibbo write of course), and the xtx not out of the sack :wink:

We can't forget Gibbo is not a real hardware reviewer, they just asked some things about benchmarking in the forum, and testing with a system previously installed with NV card.

All testing is done on both XP and Vista. The system is setup to dual boot. As we know performance differs greatly on the two OS especially for NV cards.

Surprisingly the openGL performance is menacing, which surprises me greatly, I am just having issues with D3D stuff which has got to be the drivers or due to it been an OS with an NV card in previously.

I waiting for the real reviews before i write any deep conclusion, from what Gibbo write the things looks very good for AMD.

Razor1

12-Apr-2007, 14:11

If the card can accept two 6-pin PCIe connectors then it's physically impossible for the chip, memory, cooler and every other thing on the card combined, to draw more than 225W even at the very hightest instantaneous peak of power usage. In fact, there has to be a little margin for error, so the peak usage cannot ever actually hit 225W.

At the AMD press conference a figure of approximately 200W was mentioned - I see no reason to doubt this.

well in the conference call, the 260 watt question was side stepped, the 200w was mentioned for doing GPGPU, not in game usage.

Jawed

12-Apr-2007, 14:13

At the AMD press conference a figure of approximately 200W was mentioned - I see no reason to doubt this.
But an XT with 512MB of RAM would draw less power than XTX for two reasons: GPU core speed and additional RAM.

Though GDDR4 is less power-hungry than GDDR3, so the marginal RAM power consumption might be lower than we're thinking.

But the rumour of "XTX will come later" seems ominous.

Jawed

Razor1

12-Apr-2007, 14:13

What part you write this conclusion?

They write Better IQ, more features, less price, better vista performance, fan quiet in load, for me looks like the xt will be a hot seller, and the xtx not out of the sack :wink:

We can't forget Gibbo is not a real hardware reviewer, they just asked some things about benchmarking in the forum, and testing with a system previously installed with NV card.

Wel more concrete hints about performance would be nice :) is what I'm saying, we already know its going to be around the g80 at least. And we don't know if he has a xt or xtx at least he didn't seem to say anything to that effect.

I don't know how he can say better IQ if it was glitchy, I dont' expect much difference in this departement anyways, how much better then no optimization in AF at all can you get?

Kaotik

12-Apr-2007, 14:15

well in the conference call, the 260 watt question was side stepped, the 200w was mentioned for doing GPGPU, not in game usage.

270W actually, if you watched the video rather than just read what they wrote on the site, and it wasn't the only mistake on the site.
They never answered the question, that's true, but I don't see any reason to think it would mean that it actually is 270W.
Even if Gibbo has 512MB XT model and not XTX with 1GB, there's no way the XTX would eat ~50W more

Razor1

12-Apr-2007, 14:17

270W actually, if you watched the video rather than just read what they wrote on the site, and it wasn't the only mistake on the site.
They never answered the question, that's true, but I don't see any reason to think it would mean that it actually is 270W.
Even if Gibbo has 512MB XT model and not XTX with 1GB, there's no way the XTX would eat ~50W more

yes it could each ram chip will have around 2 to 3 more watts per chip and the increased clocks of the core added to this. Well maybe not 270 but around 250 I can see it if this is the XT that Gibbo has of course.

Jawed

12-Apr-2007, 14:19

Even if Gibbo has 512MB XT model and not XTX with 1GB, there's no way the XTX would eat ~50W more
Hmm, maybe that rumour of 1GHz XTX was true, huh? :mrgreen:

Jaed

nicolasb

12-Apr-2007, 14:22

Why has everyone concluded that Gibbo has the XT version? All he's said is that it isn't the 12.4" OEM version.

I'm very encouraged by what he says about how quiet the cooler is. I'm plotting a dual-use machine that can function as a high(ish)-end rig but double up as an HTPC. R600 sounds like it'll be ideal.

_xxx_

12-Apr-2007, 14:23

Why has everyone concluded that Gibbo has the XT version?

Why has anyone concluded that he has anything at all?

vertex_shader

12-Apr-2007, 14:24

I don't know how he can say better IQ if it was glitchy, I dont' expect much difference in this departement anyways, how much better then no optimization in AF at all can you get?

32x af will be nice :wink:

Jawed

12-Apr-2007, 14:26

Why has everyone concluded that Gibbo has the XT version?
Because the rumour of XTX coming "later" won't go away.

Though it is funny that he's got this card, well before the "editors' day". AMD must know the kind of habits he has with regard to pre-release hardware. Does OcUK have much of a profile as a system builder? OK, so they're big in the UK, but that big? Perhaps for enthusiast hardware, but not as a system builder.

Jawed

Razor1

12-Apr-2007, 14:26

32x af will be nice :wink:

True! and it should be doable with the extra bandwidth, but Gibbo was testing or hopefully was testing at same settings lol.

vertex_shader

12-Apr-2007, 14:48

True! and it should be doable with the extra bandwidth, but Gibbo was testing or hopefully was testing at same settings lol.

Maybe he is testing with maximum settings, he write before r600 has more feature this is why IQ can be better :smile:
Time will tell everything, many people take too serious already what he write, and make funny conclusions on oc.uk forum :lol:

epicstruggle

12-Apr-2007, 14:56

Damn I hate this period. Info is so close to being officially released. This trickle is killing me.

Xmas

12-Apr-2007, 14:59

What? You haven't read Fudo's sentional news?
http://www.fudzilla.com/index.php?option=com_content&task=view&id=485&Itemid=1
I have no idea, how this could work. Some kind DVI-IA "integrated audio"?
It works because HDMI is backwards compatible to DVI, and the graphics card can query device capabillities. So the link will start in "DVI compatibility mode", and when a HDMI capable device is detected the graphics card will add audio output (which is interleaved in the digital stream, so no additional pins are required).

Using a DVI-I connector on the card is the best choice because it can drive DVI single and dual link, HDMI single and dual link, and analog VGA.

Unknown Soldier

12-Apr-2007, 15:15

Damn I hate this period. Info is so close to being officially released. This trickle is killing me.

wewewe.mp3

:D

US

vertex_shader

12-Apr-2007, 15:17

But if he's only got XT, not XTX then we don't know so much...

Jawed

Now Gibbo confirmed he has the XT:
This card is not the 1GB card.
Link (http://forums.overclockers.co.uk/showpost.php?p=9110900&postcount=575)

erick

12-Apr-2007, 15:29

Now Gibbo confirmed he has the XT:

Link (http://forums.overclockers.co.uk/showpost.php?p=9110900&postcount=575)

Theo's got it right then. I'm just wondering, how much is "a lot" for Gibbo: $100 bucks or $150 bucks?

trinibwoy

12-Apr-2007, 15:32

I thought we weren't expecting the 1GB card at launch any more? We've been hearing that AMD stepped back to 512MB for a while now.

But it looks like some exciting times are ahead - bad timing for me though. The last thing I should be doing these days is following 3D hardware rumours :lol:

Jawed

12-Apr-2007, 15:33

So, has AMD hobbled the driver to prevent it running 3DMk05/06? :mrgreen:

Jawed

Geo

12-Apr-2007, 15:43

I'm not a big Gibbo guy. Can someone confirm my memory that the R520 launch showed him to be full of. . .err. . .hot air?

Gelanin

12-Apr-2007, 15:58

Well, if what Giddo says is more or less correct, and what he has is a XT, then i'd say its looking fairly good for AMD/ATI.

Geo

12-Apr-2007, 16:05

Razor1

12-Apr-2007, 16:07

hehe I remember that also, there was another thing about the 7900 he mentioned can't remember exactly what but it sounded more like a sales spin.

trinibwoy

12-Apr-2007, 16:17

IbaneZ

12-Apr-2007, 16:32

I dunno you guys. The performance is very similar to GTX in benchmark and games but I think thats the most I can really say

Agreed, the GTX is fantastic and I doubt people will get the ATI card already have a GTX unless they really dislike the driver/vista issues. However for people who have waited then the R600 will no doubt prove to be a very good alternative to NV's offering.

I guess NV better better hurry up with new vista drivers, or we'll see a lot of 8800's on ebay. :)
Gibbo seems to like the new ATI drivers/CCC.

I'll probably get the R600 Pro/XL or whatever the little brother will be called.

Unknown Soldier

12-Apr-2007, 16:38

vertex_shader

12-Apr-2007, 16:38

I dunno you guys. The performance is very similar to GTX in benchmark and games but I think thats the most I can really say

The real twist is he is using factory OC'ed 8800gtx in the tests against the r600xt :smile:

Cuthalu

12-Apr-2007, 16:53

The thing is that the GDDR4 card is expected to be a little bit later than the initial X2800XT 512MB, at least that's what I remember. So if the XT goes for around 12500(as stated by sites lately) and is cheap($500) then I'd say AMD have a winner.

US

But this is GDDR4 card, and XT costing 500 is not cheap, however not expensive either. But how expensive is XTX going to be? :E

EDIT: Gibbo has again something to say:
Its now ran 3D Mark 2005 and 2006, I am pleased with the results. :)

nicolasb

12-Apr-2007, 17:20

Ah yes, a little looking around finds Gibbo reporting in July of 2005 that he had just had a meeting with ATI and they told him that R520 was 24 pipelines and a 520mhz core. That's what I was remembering. . .Maybe ATI did actually tell him that? It could have been ATI lying to Gibbo rather than Gibbo lying to us about the conversation....

Razor1

12-Apr-2007, 17:28

3dmark 06 question on a scale to 1-10; 5 being around the 8800 gtx question Gibbo says 5-6.

Arnold Beckenbauer

12-Apr-2007, 17:33

It works because HDMI is backwards compatible to DVI, and the graphics card can query device capabillities. So the link will start in "DVI compatibility mode", and when a HDMI capable device is detected the graphics card will add audio output (which is interleaved in the digital stream, so no additional pins are required).

Using a DVI-I connector on the card is the best choice because it can drive DVI single and dual link, HDMI single and dual link, and analog VGA.

Thanks. But dual link HDMI is not possible, it's always single link.

Kaotik

12-Apr-2007, 17:48

Thanks. But dual link HDMI is not possible, it's always single link.

Actually there is "Type B HDMI" which is dual-link, but it's not generally used at all, yet, it's for the future with again higher resolution displays etc.

Unknown Soldier

12-Apr-2007, 17:53

But this is GDDR4 card, and XT costing 500 is not cheap, however not expensive either. But how expensive is XTX going to be? :E

EDIT: Gibbo has again something to say:

No it's not, it's the XT card which is only 512MB DDR3.

I expect the R600 to cost $600.

US

Arnold Beckenbauer

12-Apr-2007, 17:58

Actually there is "Type B HDMI" which is dual-link, but it's not generally used at all, yet, it's for the future with again higher resolution displays etc.

Never heard about Type B. Thanks. :yes:

Sound_Card

12-Apr-2007, 18:01

Considering the possibilty of K10 and R600 co-launch, I would have to point this out...

http://www.theinquirer.net/default.aspx?article=38888

Unknown Soldier

12-Apr-2007, 18:04

Ye. saw that 10 minutes ago but still have to wonder if it will be that good.

US

Razor1

12-Apr-2007, 18:06

I wouldn't be suprised clock for clock in certain circumstances (most general usable) barcelona will be faster.

Evildeus

12-Apr-2007, 18:09

Well according to HFR:
- the X2900 XT 512 MB is priced 399$
- the X2900 XTX is still not a sure product
- the R600 is a little late mid may vs early may
- the RV630 and RV610 are late: available in june

So the simultaneous launch/availability seems quite compromise.

http://www.hardware.fr/news/lire/12-04-2007/

Kaotik

12-Apr-2007, 18:10

Galduta

12-Apr-2007, 18:15

Fudo . The r600 demo is ...Call of Juarez :mrgreen::mrgreen::mrgreen:

http://www.fudzilla.com/index.php?option=com_content&task=view&id=455&Itemid=1

http://www.fudzilla.com/images/stories/2007/April/graphics/r600_render.jpg

trinibwoy

12-Apr-2007, 18:16

Well according to HFR:
- the X2900 XT 512 MB is priced 399$

Hmmm I hope this isn't just the same rumour making the rounds. But if it is coming from multiple independent sources then it's probably true. I just don't understand how they are going to scale performance down from $400. There isn't a lot of room there.

3dilettante

12-Apr-2007, 18:18

There actually was some (SPEC) benchmarks of Barcelona a while back, and in those it was indeed faster than Core2's on most cases at least.
Of course, how that translates to realworld performance is yet to be seen

AMD's been very non-specific about what numbers and what tests they actually ran.

The only (kind of) hard numbers are SPEC rates, which are more measures of scalability. With a platform designed for quad-core, this is not much of a surprise.

What it means for gamers who have apps that don't really do two threads all that well is a good question.

jimmyjames123

12-Apr-2007, 18:21

Evildeus

12-Apr-2007, 18:22

Hmmm I hope this isn't just the same rumour making the rounds. But if it is coming for multiple independent sources then it's probably true. I just don't understand how they are going to scale performance down from $400. There isn't a lot of room there.Independant and reliable source as stated in the news. Well, i don't know, perhaps being late means being quite agressive in terms of price....

jimmyjames123

12-Apr-2007, 18:26

Being late is a bummer for AMD/ATI. If the R600 was ready a few months ago, then the X2900 512MB model probably could have sold for much higher margins with $500-600 MSRP.

trinibwoy

12-Apr-2007, 18:28

Being late is a bummer for AMD/ATI. If the R600 was ready a few months ago, then the X2900 512MB model probably could have sold for much higher margins with $500-600 MSRP.

I don't see why it can't sell for that much now if performance is there. There is still a pretty big market out there waiting to step up.

leoneazzurro

12-Apr-2007, 18:34

http://www.fx57.net/?p=576

If these tests are true, it's very strange to see XT being so much slower than XTX, maybe clocks are consistently lower.

Or, the XT could be what GTS is for GTX.

Skinner

12-Apr-2007, 18:36

Well according to HFR:
- the X2900 XTX is still not a sure product

http://www.hardware.fr/news/lire/12-04-2007/

They're kidding right?

edit;
Thanx leoneazzurro, I can sleep "again" tonight ;)

nicolasb

12-Apr-2007, 18:43

They didn't say the XTX is "not a sure product" they said it's uncertain whether it will launch at the same time as the XT.

epicstruggle

12-Apr-2007, 18:46

They didn't say the XTX is "not a sure product" they said it's uncertain whether it will launch at the same time as the XT.
I wonder if they might not just release R650 earlier to fill this gap. Arg.... too many possibilities...

Skinner

12-Apr-2007, 18:52

They didn't say the XTX is "not a sure product" they said it's uncertain whether it will launch at the same time as the XT.

That sounds better to me :)

icecold1983

12-Apr-2007, 18:54

so now the rumors are that it wont be a 'unified launch'? this just gets better and better.

suryad

12-Apr-2007, 18:55

This sort of wait for the R600 is reminding me of when the first star wars episode.... meaning episode 1 was coming out....all the anticipation.

Geeforcer

12-Apr-2007, 19:13

Ye, i'm beginning to think the real reason why AMD never launched the R600 in March was because Dell(and other OEM's) bought all the chips and placed orders for the first batch of RV610/630's and thus AMD has now supplied them or is still delivering on the 'Million plus chips' (http://www.fudzilla.com/index.php?option=com_content&task=view&id=499&Itemid=34) that is supposed to be delivered and not the original rumours that floated about saying that the chip/card had problems.

My memory is not what it used to be, but not only would such delay be a first, it would also not explain why R600 was delayed if OEMs ordered all those RVxxx chips.

Hmm.. now i'm thinking that AMD might actually have done something right in order to keep the company above the water. For us consumers though it's a bitter pill but I guess that's where the money lies.

So AMD made more money by not selling chips? And OEMs don't care that they are getting what amounts to second-rate product (R600@80nm)? Again, I don't see how this got AMD ahead, financially.

Now i'm beginning to think that AMD might actually release the 65nm R650 for retail consumers at release date instead of the 80nm R600 which they'll sell to the OEM's before hand.

Personally that does not seem likely to me, but we'll see.

PatrickL

12-Apr-2007, 19:14

http://www.fx57.net/?p=576

If these tests are true, it's very strange to see XT being so much slower than XTX, maybe clocks are consistently lower.

Or, the XT could be what GTS is for GTX.

Catalyst 7.1 ? Seems strange as cards should be tested with 7.4 or so by now no ?

flippin_waffles

12-Apr-2007, 19:28

Catalyst 7.1 ? Seems strange as cards should be tested with 7.4 or so by now no ?

Would the older catalyst drivers support the next gen cards? I have no idea of course.
As for Gibbo's card, I wonder why they would send him alpha/beta drivers. Maybe they are playing the game that NV started with driver exploitation. There seems to be a small consensus that NV might be holding back some performance through drivers to throw AMD off with performance potential. Sounds reasonable to me, and would make sense why they aren't shipping him WHQL drivers, as AMD has said they're drivers are ready to go.
And with all the misinformation floating around, it would make sense that AMD is playing possum, as in "two can play that game, NV". I just have a sneaking suspision that that is what's going on here.

Yeah I know, that redefines "stretching it", but reasonable, no?? :lol: :razz: :oops:

Evildeus

12-Apr-2007, 19:30

They didn't say the XTX is "not a sure product" they said it's uncertain whether it will launch at the same time as the XT.
Yeah sorry about that :oops:

Kaotik

12-Apr-2007, 19:37

Would the older catalyst drivers support the next gen cards? I have no idea of course.
As for Gibbo's card, I wonder why they would send him alpha/beta drivers. Maybe they are playing the game that NV started with driver exploitation. There seems to be a small consensus that NV might be holding back some performance through drivers to throw AMD off with performance potential. Sounds reasonable to me, and would make sense why they aren't shipping him WHQL drivers, as AMD has said they're drivers are ready to go.
And with all the misinformation floating around, it would make sense that AMD is playing possum, as in "two can play that game, NV". I just have a sneaking suspision that that is what's going on here.

Yeah I know, that redefines "stretching it", but reasonable, no?? :lol: :razz: :oops:

At least none of the Vista Catalysts so far has supported R600, of course it's possible they have separate R600 drivers with same names, but I kinda doubt they're using the same names for them before the support for them is included in the releases.

jimmyjames123

12-Apr-2007, 20:04

I don't see why it can't sell for that much now if performance is there. There is still a pretty big market out there waiting to step up.

Let's suppose that AMD did sell the X2900XT for $500-$600 in May (?) of this year when it is launched. By then, in that price range, it would almost certainly be competing against a beefed up 8800 high end variant in that price range. If it really was as fast or faster than anything else NV will have to offer in that range, then sure they could sell it for that. However, if it's closer to the current 8800 GTX with respect to performance, then they would have to price it much lower than that, whereas if it was released a few months ago they could have priced it at or above 8800 GTX levels ($649 MSRP at launch IIRC).

zealotonous

12-Apr-2007, 20:22

I could see an X2900 with 512MB RAM having a list price of around $400. After all, an 8800 GTS with 640MB RAM had a list price of $449, and is now selling for probably $100 lower than that. Especially at that price, it would be a really great value.

One advantage that ATI has in going with the 512bit bus is that they can save money on RAM while still having massive bandwith. It won't be easy for NV to counter until they move up to a 512bit bus, unless they have some other trick up their sleeve.

I am not sure if there is an actual advantage to ATI. The 512 bit bus is not an advantage to the OEM as they have to add quite a bit more traces to the PCB thus costing more to the end user or lowering margins to the OEM. Does the cost of having a more complex PCB outweigh the costs of more ram? Perhaps a performance comparison would be helpful once the cards are released. I would think it would vary depending on the application and the bandwidth needs. Regardless, it would be an interesting comparison.

trinibwoy

12-Apr-2007, 20:29

Let's suppose that AMD did sell the X2900XT for $500-$600 in May (?) of this year when it is launched. By then, in that price range, it would almost certainly be competing against a beefed up 8800 high end variant in that price range. If it really was as fast or faster than anything else NV will have to offer in that range, then sure they could sell it for that. However, if it's closer to the current 8800 GTX with respect to performance, then they would have to price it much lower than that, whereas if it was released a few months ago they could have priced it at or above 8800 GTX levels ($649 MSRP at launch IIRC).

Correct but given the little tidbits leaking thus far I'm leaning towards AMD not having any trouble competing with a G80 refresh. It is pretty much confirmed that there are 16 2200Mhz memory modules on the XTX. That's 140GB/s of bandwidth right there. I know that doesn't tell the whole story but it leads me to think that R600 can make use of all that bandwidth. If it can't why would AMD waste money on the fastest GDDR4 available?

Razor1

12-Apr-2007, 20:46

I am not sure if there is an actual advantage to ATI. The 512 bit bus is not an advantage to the OEM as they have to add quite a bit more traces to the PCB thus costing more to the end user or lowering margins to the OEM. Does the cost of having a more complex PCB outweigh the costs of more ram? Perhaps a performance comparison would be helpful once the cards are released. I would think it would vary depending on the application and the bandwidth needs. Regardless, it would be an interesting comparison.

OEM's don't build the cards, they just buy the card and use it in thier systems. ie: Like OEM mice, or keyboards.

Kaotik

12-Apr-2007, 20:47

Correct but given the little tidbits leaking thus far I'm leaning towards AMD not having any trouble competing with a G80 refresh. It is pretty much confirmed that there are 16 2200Mhz memory modules on the XTX. That's 140GB/s of bandwidth right there. I know that doesn't tell the whole story but it leads me to think that R600 can make use of all that bandwidth. If it can't why would AMD waste money on the fastest GDDR4 available?

It's not pretty much confirmed, but 100% confirmed that at least part of OEM models have those.

erick

12-Apr-2007, 20:47

Correct but given the little tidbits leaking thus far I'm leaning towards AMD not having any trouble competing with a G80 refresh. It is pretty much confirmed that there are 16 2200Mhz memory modules on the XTX. That's 140GB/s of bandwidth right there. I know that doesn't tell the whole story but it leads me to think that R600 can make use of all that bandwidth. If it can't why would AMD waste money on the fastest GDDR4 available?

Because then they would need to squeeze every drop of performance out of the card by any means necessary? ;)

By the way, just visited Samsung's website today, and I think I saw 512Mbit GDDR4 2800MHz modules available. So 2200MHz may not be the fastest there is.

trinibwoy

12-Apr-2007, 20:57

By the way, just visited Samsung's website today, and I think I saw 512Mbit GDDR4 2800MHz modules available. So 2200MHz may not be the fastest there is.

Yeah you're right 2200Mhz is the lowest grade of GDDR4 they offer. I misread that table earlier. But that doesnt minimize the shear amount of bandwidth it will have available. The XT will have a big advantage over the GTS there too (assuming they go head to head) - with 1600Mhz GDDR3 it will have 60% higher bandwidth.

Sound_Card

12-Apr-2007, 21:08

an XT for $400 seems very possible to me. Plenty of room to work with IMO. This should make a xtx priced at $500 and if the performance numbers are true, their is plenty of room in both performance and price for a stop gap card between the xt and xtx. Like for instance, a x2950pro later down the line.

R600

XTX = $500
XT = $400
XL = $300 - $350
gto = $250

RV630

XT = $200
pro = $150

RV610

XT = $100 - $125
pro = $50 - $75

Then if the R650 roumors are ture, that being coinciding with R600 with in a month or two, then their is your $600 price point card.

turtle

12-Apr-2007, 21:53

Ollo

12-Apr-2007, 21:57

So this is page 79 of the second or third R600 rumours thread. We have all been eyeing this thread a couple of times every day. You know what the worst thing about that is? Knowing that Dave B is lurking...

Jawed

12-Apr-2007, 22:12

Surely 8800GTX will hit whatever price point it needs to, to compete with HD2900XT. $400, $350, whatever.

That'll also push 8800GTS down into the $200-250 bracket, so that those disappointed with 8600GTS will have a nice alternative.

Jawed

Unknown Soldier

12-Apr-2007, 22:15

My memory is not what it used to be, but not only would such delay be a first, it would also not explain why R600 was delayed if OEMs ordered all those RVxxx chips.

You seem to forget that OEM's have also bought R600's (http://www.fudzilla.com/index.php?option=com_content&task=view&id=484&Itemid=1). That's a OEM R600 pictured if I remember correctly since the OEM version is 12.4" and the Retail is 9.5".

So AMD made more money by not selling chips? And OEMs don't care that they are getting what amounts to second-rate product R600@80nm? Again, I don't see how this got AMD ahead, financially.

Again, you think that OEM's can just sit around and wait for 65nm parts? (http://www.fudzilla.com/index.php?option=com_content&task=view&id=326&Itemid=1) If they going after 80nm RV6xx parts, then they'll certainly go after the 80nm R600. The 12.4" part proves this.

Personally that does not seem likely to me, but we'll see.

Yep, that's what we doing.

US

Unknown Soldier

12-Apr-2007, 22:20

Knowing that Dave B is lurking...

Ye, I think we need to bring out the Psycho music again. :D

US

Razor1

12-Apr-2007, 23:22

You seem to forget that OEM's have also bought R600's (http://www.fudzilla.com/index.php?option=com_content&task=view&id=484&Itemid=1). That's a OEM R600 pictured if I remember correctly since the OEM version is 12.4" and the Retail is 9.5".

Again, you think that OEM's can just sit around and wait for 65nm parts? (http://www.fudzilla.com/index.php?option=com_content&task=view&id=326&Itemid=1) If they going after 80nm RV6xx parts, then they'll certainly go after the 80nm R600. The 12.4" part proves this.

Yep, that's what we doing.

US

There might be some truth to this, but in the volume they are talking about, I'm a bit weary talked with a few guys at Dell and they hinted at they their partnership with AMD is strong, and getting stronger on the cpu front (possibly hinting at more AMD boxes coming out), but nothing out of the ordinary, to me it sounded like they are buying ATi cards as they did before, but allocation seems to be the same as they were in the past.

Unknown Soldier

12-Apr-2007, 23:30

Luckily, Dell ain't the only OEM's. :)

US

Razor1

12-Apr-2007, 23:33

thats true, but they are one of the biggest OEM's AMD has if not the biggest lol, cause HP is almost exclusively nV for the past year, if not only nV.

INKster

12-Apr-2007, 23:38

thats true, but they are one of the biggest OEM's AMD has if not the biggest lol, cause HP is almost exclusively nV for the past year, if not only nV.

Dell buys a lot from Nvidia too, let us not forget that.

Kaotik

12-Apr-2007, 23:46

If the drivers are in as good shape as at least I've understood they should be, AMD should see quite big wins on the OEM sector due nVidias driver situation, not limited to just Delll.

Frank

12-Apr-2007, 23:48

So, in short: No, the R600 isn't going to beat the 8800. As expected.

Although both will have their own strengths and weaknesses.

Razor1

12-Apr-2007, 23:51

well the Ultra is coming and I think that $1000 price is there..........

Kaotik

13-Apr-2007, 00:03

So, in short: No, the R600 isn't going to beat the 8800. As expected.

Although both will have their own strengths and weaknesses.

At what point did we come to this conclusion? :shock:
There has been more and more suggestions, with some proper arguments behind these suggestions, that Gibbo or whatever his name was doesn't actually have R600 at all.

INKster

13-Apr-2007, 00:09

So, in short: No, the R600 isn't going to beat the 8800. As expected.

Although both will have their own strengths and weaknesses.

I've seen nothing stating that yet.
Perhaps they don't beat it in the time-to-market angle, but that should be it. Performance prospects look good so far.

Geo

13-Apr-2007, 00:15

Maybe ATI did actually tell him that? It could have been ATI lying to Gibbo rather than Gibbo lying to us about the conversation....

He could secretly be Dave Orton's love child too. But he's not.

I'm not going to get into a whole long historical digression on that topic, but for my purposes I'm quite satisfied that makes him a knowing liar in the summer of 2005 rather than just an enthusiast who speculated wrongly. YMMV. And, y'know, at the end of the day what he was doing then may not be what he's doing now. But personally I'm not prepared to trust it.

Shtal

13-Apr-2007, 01:58

So, in short: No, the R600 isn't going to beat the 8800. As expected.

Although both will have their own strengths and weaknesses.

I wonder how R600 512Bit memory vs. 384Bit memory G80 will perform @ super AA and including high resolution.
Probably resemble - similar comparison like 256Bit Radeon 9700Pro vs. 128bit Geforce 4 Ti4600. :)

Pete

13-Apr-2007, 02:05

At some point, casting for analogies to past generations ceases to be enlightening. I think we've long past that point.

Feh. Not much more tortuous suspense to endure, at least.

Shtal

13-Apr-2007, 02:15

At some point, casting for analogies to past generations ceases to be enlightening. I think we've long past that point.

Feh. Not much more tortuous suspense to endure, at least.

At your wise knowledge would you consider logically 100GB bandwidth could be eaten up near future with high demanding graphical games using extreme detail settings.

INKster

13-Apr-2007, 02:22

At your wise knowledge would you consider logically 100GB bandwidth could be eaten up near future with high demanding graphical games using extreme detail settings.

Extreme detail may also mean higher resolution textures in future games, in which case the amount of texture memory becomes even more relevant. That's why the 1GB R600 would have been useful to have a proper comparison at that level.
And since this first "R600 with GDDR3" appears to have "only" a 512MB frame buffer, versus the 768MB of a standard 8800 GTX, the bandwidth advantage may not be so overwhelming in actual game play.
G80 is not exactly starving for bandwidth as of yet.

Shtal

13-Apr-2007, 02:30

Extreme detail may also mean higher resolution textures in future games, in which case the amount of texture memory becomes even more relevant. That's why the 1GB R600 would have been useful to have a proper comparison at that level.
And since this first "R600 with GDDR3" appears to have "only" a 512MB frame buffer, versus the 768MB of a standard 8800 GTX, the bandwidth advantage may not be so overwhelming in actual game play.
G80 is not exactly starving for bandwidth as of yet.

I was actually talking about 1GB Memory R600 512Bit.
But I see your point! :)

Pete

13-Apr-2007, 03:22

At your wise knowledge would you consider logically 100GB bandwidth could be eaten up near future with high demanding graphical games using extreme detail settings.
I am neither wise nor knowledgeable (particularly about R600), but hasn't anyone benchmarked an 8800 with both stock and overclocked RAM? Otherwise, all I've got to go on are sites (http://www.tweaktown.com/articles/983/9) that have OCed the RAM along with the core, which isn't entirely useful when trying to determine bandwidth sensitivity. OCed 8800GTXs are hitting 100GB/s with 2GHz RAM, so someone with a card can definitely answer your question.

---

As for R600XTX's purported 12k 3DM06 score, an upclocked 8800GTX seems (http://www.tweaktown.com/articles/983/3/page_3_benchmarks_3dmark06/index.html) capable (http://www.nordichardware.com/Reviews/?page=4&skrivelse=493) of reaching that #, so that figure's not super-interesting in terms of fleshing out R600's strong points. Still, nice to know it'll at least be competitive.

icecold1983

13-Apr-2007, 03:24

8800gtx only seems to rly benefit from ocing the core. so its probably texture/fillrate/rop limited. not that surprising since with a 1.35 ghz shader core its unlikely any current games are gonna bottleneck it.

INKster

13-Apr-2007, 03:34

DT has the scoop:

http://www.dailytech.com/article.aspx?newsid=6903

320 SP's, 24x AA, 128bit HDR, among other juicy stuff.

Geo

13-Apr-2007, 03:38

DT has the scoop:

http://www.dailytech.com/article.aspx?newsid=6903

320 SP's, 24x AA, 128bit HDR, among other juicy stuff.

24xAA? Hmm. Everybody clap for Tinkerbell. . . .

INKster

13-Apr-2007, 03:39

24xAA? Hmm. Everybody clap for Tinkerbell. . . .

:D You noticed it too, hey ?

Geo

13-Apr-2007, 03:41

SuperAA maybe?

I mean, believe me, I'll be the first to line up with a "Right on!" if they are doing that with a single gpu. :smile: And, y'know, it wouldn't surprise me if they had a goodie in their pocket they've been hanging on to for the last minute.

R300King!

13-Apr-2007, 04:03

24X AA? I hope this info is correct. It should be or Daily Tech's neck will be severed. lol

Maybe some of these game Gibbo has been testing has the AA settings at max, etc and so the HD2900XT's score is low or on par with the GTX but he is not really paying attn to the quality of the picture. Maybe the 2900's IQ is better but at the same FPS. Gibbo did say the 2900's image did look better and more vivid.

jimmyjames123

13-Apr-2007, 04:12

Correct but given the little tidbits leaking thus far I'm leaning towards AMD not having any trouble competing with a G80 refresh. It is pretty much confirmed that there are 16 2200Mhz memory modules on the XTX. That's 140GB/s of bandwidth right there. I know that doesn't tell the whole story but it leads me to think that R600 can make use of all that bandwidth. If it can't why would AMD waste money on the fastest GDDR4 available?

I don't doubt that AMD can compete with an XTX with that much memory, I am just saying that they missed out on an opportunity to get really high profit margins when their product was delayed and set to launch about 6 months after the 8800 GTX.