Official HD 2900 Discussion Thread

[onethreehill]

And will aftermarket coolers (core only) for teh GTX work on the 2900?

And when the hell can we expect the XTX?

Maybe Q3

[rob[GL]]

I thought the ATi CC can do it.

EDIT: I think people also use this to monitor temps.

http://forums.techpowerup.com/showthread.php?t=30555

I don't have 8-pin PCI-Express, so I can't get the overdrive tab, which shows temps...

[dietwaterrr]

I know, but if the tradeoff is going with a worse mobo for the new ATi cards, I would reconsider. Which do you think is better?

you said u would trade ur gtx for a pair of 2900xt. you cant crossfire on a nvidia chipset.

[rob[GL]]

Okay, so temperatures as requested;

Idle: 60c
Load: 70c

[xVeinx]

thats not bad at all...

[CedricFP]

70 load isn't bad at all!

I'm getting more and more anxious by the minute. In fact, I'm heading out now to see if I can get one of these bad boys... then my rig will be complete!

[[XC] Lead Head]

@Rob GL,

Nice performance, but is that Half Life 2: Episode Two or 1?

[theteamaqua]

the 4 screws next to the core looks about the same size as X1900 ... maybe D Tek Fuzion GPU will do the trick ...

[kuhla]

@Rob GL,

Nice performance, but is that Half Life 2: Episode Two or 1?

Its the "Black Box" deal so you get....

Half-Life 2: Episode 2
Portal
Team Fortress 2

EDIT : I should probably mention these games aren't out yet. What you get is a voucher to download it for free on steam when it is released (right?).

.....and its supposed to retail for $40 so that is added value.

I ran some calculations at 1920x1200 from anandtech's graphs and I found that the 2900XT is on average 2 FPS faster than the GTS640 and if we were to make that into a % average that would come out to the 2900XT is on average 6% faster than the GTS640. For ~$100 more that seems a bit steep

[theteamaqua]

GTS 640 is about low $300's now .... EVGA $329.99 after rebate ... i bought my 2nd one almost 3 months ago and cost 10 bux more ...

[Caparroz]

I think a lot of people here are in denial and I kind of understand that. Many of us just waited an waited for something that is, let's face it, a major disappointment. The hype over R600 was something downright insane. Everybody was like "R600 wil be at least 50% faster then 8800gtx and with a much better IQ on top of that".

A few of weeks ago when it became evident that HD2900xt wouldn't be the G80 killer that everyone took for granted, some argued that future drivers would solve that. More in-depth reviews like the ones at Anand and Beyond3D clearly state that R600 architecture is extremely software dependant:

If it seems like all this reads in a very complicated way, don't worry: it is complex. While AMD has gone to great lengths to build hardware that can efficiently handle parallel data, dependencies pose a problem to realizing peak performance. The compiler might not be able to extract five operations for every VLIW instruction. In the worst case scenario, we could effectively see only one SP per block operating with only four VLIW instructions being issued. This drops our potential operations per clock rate down from 320 at peak to only 64.

But maximizing throughput on the AMD hardware will be much more difficult, and we won't always see peak performance from real code. On the best case level, R600 is able to do 2.5x the work of G80 per clock (320 operations on R600 and 128 on G80). Worst case for code dependency on both architectures gives the G80 a 2x advantage over R600 per clock (64 operations on R600 with 128 on G80).

While NVIDIA focused on maximizing parallelism in this area of graphics, AMD decided to try to extract parallelism inside the instruction stream by using a VLIW approach. AMD's average case will be different depending on the code running, though so many operations are vector based, high utilization can generally be expected.

While going 5-way scalar has allowed AMD more flexibility in instruction scheduling compared to their previous hardware, that flexibility arguably makes your compiler harder to write, not easier. So as a driver writer you have more packing opportunities -- and I like to think of it almost like a game of Tetris when it comes to a GPU, but only with the thin blocks and with those being variable lengths, and you can sometimes break them up! -- those opportunities need handling in code and your corner cases get harder to find.

The end result here is a shader core with fairly monstrous peak floating point numbers, by virtue of the unit count in R600, its core clock and the register file of doom, but one where software will have a harder time driving it close to peak. That's not to say it's impossible, and indeed we've managed to write in-house shaders, short and long and with mixtures of channels, register counts and what have you, that run close to max theoretical thoughput. However it's a more difficult proposition for the driver tech team to take care of over the lifetime of the architecture, we argue, than their previous architecture.

But what people don't realize is that ATI had a billion years to develop proper drivers. It doesn't matter what ATI says. The R600 is very late to the game, unfortunately. That "magic driver" is to be taken with a HUGE grain of salt. The IQ loss is very evident.

Like I said in another 2900xt thread:

"the HD2900xt is such a disaster that instead of bringing Nvidia's prices down, it made'em go up. Yesterday the cheapest 8800gts 320 was $250 and now is $295 at newegg:

http://www.newegg.com/Product/Produc...613&name=320MB

Before anyone call me a fanboy let me put this: I owned only ATI cards since R300. Not beacause I like ATI and hate Nvidia but just cuz they were better and faster. I too had hope that R600 would be a beast of a GPU but, unfortunately, it isn't. In my book it is a major flop and I truly hope that AMD/ATI get their act together and put some serious competition to both Nvidia and Intel

[AbelJemka]

http://www.hardware.fr/articles/671-...d-2900-xt.html
French test better driver. on Vista.





All is not perfect but seeing this review and Digit-Life one after one day is begin to be very interesting.

@Caparroz : Using Anandetch review is good but u can use it entirely. In their image quality part they say nothing wrong. Image Quality in Hardware.fr (BeHardware) review is equal to X1900 image quality but a little inferior to 8800 image quality.

[Caparroz]

^^I was talking about the Call of Juarez test. The 2900xt fireplace pic clearly shows blurred textures. That bench used the "magic" 8.37.4.2 driver, didn't it?

[theteamaqua]

]
Before anyone call me a fanboy let me put this: I owned only ATI cards since R300. Not beacause I like ATI and hate Nvidia but just cuz they were better and faster. I too had hope that R600 would be a beast of a GPU but, unfortunately, it isn't. In my book it is a major flop and I truly hope that AMD/ATI get their act together and put some serious competition to both Nvidia and Intel

yes i too hoped that R600 was gonna be fast ... so that nv will drop price and consumer is the winner, but no ... nv didnt drop

i see a lot of peopel who hyped this card tried damage control ...

heres what most of em said :
1. HD2900XT was only meant to compete with GTS

2. image quality is better

3. immature dirver

i will say that for $399 its a helluva card , and from what i have seen its a good 3dmarkcard .... i'd say it score the same as a GTX ie 2900XT 850/2000 vs GTX @ 660/2100, but games .... GTX wins hands down

[ownage]

PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE PLEASE

be true

http://www.computerbase.de/news/trei...ibervergleich/
There's the link

[Ubermann]

Thats some sick improvment, can anyone get that driver ?
But with the loss of quality =(

[Shintai]

Its easy to enhance performance if you can sacrifice other things...

[AbelJemka]

All modern test have IQ part, so cheat with IQ is the best way to be caught. May be demo bug or drivers issues but for some it seems esier to call it cheating.

[Eastcoasthandle]

I think a lot of people here are in denial and I kind of understand that. Many of us just waited an waited for something that is, let's face it, a major disappointment. The hype over R600 was something downright insane. Everybody was like "R600 wil be at least 50% faster then 8800gtx and with a much better IQ on top of that".

A few of weeks ago when it became evident that HD2900xt wouldn't be the G80 killer that everyone took for granted, some argued that future drivers would solve that. More in-depth reviews like the ones at Anand and Beyond3D clearly state that R600 architecture is extremely software dependant:



But what people don't realize is that ATI had a billion years to develop proper drivers. It doesn't matter what ATI says. The R600 is very late to the game, unfortunately. That "magic driver" is to be taken with a HUGE grain of salt. The IQ loss is very evident.

Like I said in another 2900xt thread:

"the HD2900xt is such a disaster that instead of bringing Nvidia's prices down, it made'em go up. Yesterday the cheapest 8800gts 320 was $250 and now is $295 at newegg:

http://www.newegg.com/Product/Produc...613&name=320MB

Before anyone call me a fanboy let me put this: I owned only ATI cards since R300. Not beacause I like ATI and hate Nvidia but just cuz they were better and faster. I too had hope that R600 would be a beast of a GPU but, unfortunately, it isn't. In my book it is a major flop and I truly hope that AMD/ATI get their act together and put some serious competition to both Nvidia and Intel

^^I was talking about the Call of Juarez test. The 2900xt fireplace pic clearly shows blurred textures. That bench used the "magic" 8.37.4.2 driver, didn't it?

Would you give it a rest already, that's the most broken record statement(s) of the year. Do you need the approval of others to validate your own opinion of the R600 before mature drivers are released? Also, your statement is purely one from the Nvidia fan camp. Besides, what IQ test? Do you have a link?

[Noobie]

Would you give it a rest already, that's the most broken record statement(s) of the year. Do you need the approval of others to validate your own opinion of the R600 before mature drivers are released? Also, your statement is purely one from the Nvidia fan camp. Besides, what IQ test? Do you have a link?

I have no clue what you are talking about, so I assume you mean the Call Of Juarez test.

Non-mouse over:
http://tertsi.users.daug.net/temp/R600/iq/coj_ati.jpg
http://tertsi.users.daug.net/temp/R600/iq/coj_nv.jpg

Mouse-over:
http://tertsi.users.daug.net/temp/R6...ia_vs_ati.html

Article:
http://www.legitreviews.com/article/504/2/

The "mature drivers" thing is getting annoying. How do you define drivers to be mature? When ATI releases an official version? Why am I even asking; no one is going to agree on a definition of mature....

[Eastcoasthandle]

I have no clue what you are talking about, so I assume you mean the Call Of Juarez test.

Non-mouse over:
http://tertsi.users.daug.net/temp/R600/iq/coj_ati.jpg
http://tertsi.users.daug.net/temp/R600/iq/coj_nv.jpg

Mouse-over:
http://tertsi.users.daug.net/temp/R6...ia_vs_ati.html

Article:
http://www.legitreviews.com/article/504/2/

The "mature drivers" thing is getting annoying. How do you define drivers to be mature? When ATI releases an official version? Why am I even asking; no one is going to agree on a definition of mature....

Don't start that crap with me. Current drivers don't show the potential of the card yet, bottom line. The term "mature" is not relative. It has a very specific meaning. Therefore, know what you are talking about before you start blathering.
As for the photos:
-He clearly states he has pre-alpha (but doesn't mention which version). Again, drivers are NOT MATURE!
-The image test looks like the bitmap was set to quality and not High Quality. I find myself having to change this with each driver update.
-He clearly states:

When it comes to image quality we installed both drivers and without adjusting any of the settings in the control panel jumped right into the benchmark to see how they did.

But never shows a pic of those settings and never explains what those settings were at the time he provided those photos. Not only do we not know what those adjustments were between R600/G80, we have no control example to bench against at other bitmap/AA/AF settings.

In all, this is a pure example of FUD through photos. There is no supporting documentation (regarding those photos) to provide an explain of how he arrived to that conclusion (through photos). People taking these photos at face value without asking questions first is the real problem here. Why shouldn't you ask questions? It's a freaking 3 page (short) review of the HD 2900 XT.

[evolution21]

I'll wait for overclocked partner cards and Catalyst 8.38 before making a judgement. I'm thinking of buying the HD2900XT, it has about the same performance as the GTS and still has headroom for driver improvement. So i guess this should be a better card to buy than the GTS. This card shows potential, it is not as good as the GTX, but can come close to it with driver improvement (I hope).
I only had NVidia cards (GeForce 2 MX400, GeForce 5600XT and GeForce 7300LE), but now i'm thinking of changing to AMD.
I waited for this card 5 or 6 months and i hoped for alot more, but it is not that bad for the money it costs.
As for the reviews, some sound fishy to me. I don't know what to believe anymore tbh.

[awdrifter]

Any reviews of the HD2600XT? How does it compre to the X1950Pro?

[Shintai]

Any reviews of the HD2600XT? How does it compre to the X1950Pro?

Dont expect anything but around 1950pro performance +DX10. 8600GTS didn´t show wonders either. And it will have the same 128bit bus as the 8600GTS, while the x1950Pro got 256.

[bingo13]

email w0mbat

It's 8.37.4.2 the driver you want. There should be a 8.39 also but I am not sure if it's been released yet.

The 8.37.4.2 drivers are alpha and have numerous issues. They do give a glimpse as to how ATI/AMD will optimize this card. The 8.38b2 are based on this driver and offer further refinements but CrossFire/OpenGL is not working right under Vista, AVIVO is not working correctly with Blu-ray/HD-DVD, IQ is not up to par in several areas. Unless you need them to show to some nice 3DMark marks then I would wait on the next release that is scheduled on 5/23. 8.39s are in alpha testing now.