I also heard ATI eats babies!

Decent is not enough to me.
And what you are calling decent I call it poor.

From what we have found during benchmarking the ATI Radeon HD 2600 graphics card using newer drivers, only Half-Life 2, Company of Heroes and Command and Conquer: Tiberium Wars have more or less substantial performance increases. This does not dramatically change our conclusion regarding the product.

thats not the point anton......

you say you use 7.9 but in fact you probably use 7.5 or something,that's just lying to your readers,its not telling the truth.....
its just not fair....

We used Catalyst 7.9.

1. For your information: Catalyst 7.5 does not support Radeon HD 2600-series. Only special version of 7.6 and 7.7 do.

2. Again, from what we have found during benchmarking the ATI Radeon HD 2600 graphics card using newer drivers, only Half-Life 2, Company of Heroes and Command and Conquer: Tiberium Wars have more or less substantial performance increases.

3. Given that you are not quite confident in terms of drivers and performance, I am afraid that your criticism hardly makes sense since it has no base...

you used the new drivers?

so this means you also used them for the reference 2600XT and pro scores?
kinda strange you have the exact FPS numbers correct to the digit.....
even when i bench with same drivers,results usually change at least 0.1 FPS
so stop lying about it,dont start making excuses
just get it right......

From what we have found during benchmarking the ATI Radeon HD 2600 graphics card using newer drivers, only Half-Life 2, Company of Heroes and Command and Conquer: Tiberium Wars have more or less substantial performance increases. All the other results were similar. Given that, we do not perform across-the-board retesting for every review, but reuse results.

so you didnt use 7.9 catalyst as stated......
so you just said you were lying in the review....
how lame can you be
and how do you know it only improves those games? you guys didnt tested it
LOL

Can you read? Or, can you understand what you are reading?

We used Catalyst 7.9 for testing the ATI Radeon HD 2600 Pro from Power Color.

From what we have found during benchmarking the ATI Radeon HD 2600 graphics card using newer drivers, only Half-Life 2, Company of Heroes and Command and Conquer: Tiberium Wars have more or less substantial performance increases. All the other results are similar across different driver versions for HD 2600-series. Given that, we do not perform across-the-board retesting for every review, but reuse results.

i can read
so just state in the review that you also use older drivers for reference results.....
if you just say 7.9,you are kind of lying to your readers...

If you can read - reread my posts then...

From what we have found during benchmarking the ATI Radeon HD 2600 graphics card using newer drivers, only Half-Life 2, Company of Heroes and Command and Conquer: Tiberium Wars have more or less substantial performance increases. This does not dramatically change our conclusion regarding the product.

Given that in the last two years we tested a number of performance-mainstream graphics cards with 256-bit memory bus and decent performance with FSAA 4x, we have no idea why we have to step back to no-FSAA mode ahead of the ATI RV670 and Nvidia G8M launch, both of which are projected to feature 256-bit memory bus.

Please keep in mind that enabling FSAA does not affect relative performance of different products.

If Radeon HD 2600 Pro cannot enable FSAA in 3-years old Far Cry game, what this card is for then?

“If Radeon HD 2600 Pro cannot enable FSAA in 3-years old Far Cry game, what this card is for then?”

Maybe DX10?
Maybe lower power consuming?
Maybe the HD Video processing engine?
Maybe because it's cheap to produce and cost much less to us?

I don’t see any disadvantage. Sure It could perform a little better, specially with AA. But guess what; most of my friends doesn’t know what’s that and where they enable it, in the end useless feature.

1. Are you capable of playing FPS/RTS/TPS games at 15 frames per second? For DirectX 10 performance test please see a separate article:
http://www.xbitlabs.com/articles/video/display/directx10-game s_9.html#sect0

2. Lower power consumption compared to what?

ATI Radeon HD 2600 XT - 28/48 (peak 2D/peak 3D)
ATI Radeon HD 2600 Pro - 21/30
Nvidia GeForce 6800 (NV42) - 31/38
Nvidia GeForce 6600 GT - 39/47

Sources:

http://www.xbitlabs.com/articles/video/display/gpu-consumptio n2006_7.html
http://www.xbitlabs.com/articles/video/display/powercolor-hd 2600pro_4.html#sect0

Just for information: it is possible to cool-down graphics cards with about 60W TDP using passive cooling system.

It is interesting to note how those graphics cards perform in Far Cry, Pier level, FSAA 4x with 16x AF, 1280x1024:

ATI Radeon HD 2600 XT - 21
ATI Radeon HD 2600 Pro - 14.9
Nvidia GeForce 6800 (NV42) - 40
Nvidia GeForce 6600 GT - 33.5

Sources:

http://www.xbitlabs.com/articles/video/display/msi-nx6800-sli _9.html
http://www.xbitlabs.com/articles/video/display/powercolor-hd 2600pro_8.html#sect0

So, the Radeon HD 2600 XT consumes the same amount of power compared to GeForce 6800/NV42, while performing two times faster in the same condition. Is it good?

3. HD video post-processing engine is an indisputable advantage of the ATI Radeon HD 2600. However, if you are talking about low CPU utilization, then this advantage is hardly important. BTW. Do you already have a BD or HD DVD drive inside your PC?

4. Do you known how cheap the ATI Radeon X1950 Pro and Nvidia GeForce 6800/NV42 are? :))

5. If you/your friends do not know what is FSAA and where they enable it, then they hardly need detailed review, but should be happy with 3Dmark scores, which we publish without FSAA.

Anton thanks for the reply.

1. Yes I have to play because of lame programmers! But I can also lower detail settings. Right?
On the same test:
http://www.xbitlabs.com/articles/video/display/directx10-game s_6.html
Ultra high end video cards that cost 400$ are doing 25fps. Now I return you question to you, how bad are 15fps in 100$ card?


2. Which of all those cards is lacking features?
Want to bet that without AA all cards are more playable and the new Ati cards win most of the tests and still cost less the the other much older cards?

And if in the charts the power consuming numbers of the 1950Pro for example seams low, why did most of my friends had to swap the 400W PSU for one 600W, any other PSU would shutdown their computers with that card while playing, want to bet with one 2600 that wouldn’t happen?
Just go to the forums and search for the sapphire 1950Pro and the PSU problems.

And you example is very bad, why not search for one which is optimized for Ati like the link you posted in (1)?


3. It’s not important are you kidding? The CPU sits idle doing practically nothing and that’s not important? Tell me you think the Ati cards consume too much for the performance they delivers. Now they can do calcs that avoids CPU work (CPU power consuming) so what’s your point after all? Also isn’t Xbit one of those sites that bash Ati and Nvidia high end card for the lack of Video processor?


4. I do know. But I also know they lack DX10, they need a much powerfull PSU to work, and that they still cost more. Also some games require SM3 model to work, how much time will take for the need of SM4 a requirement to play the game?


5. 3Dmark is not a computer game. I and my friend cant base the video card choice in one stupid test.
Xbit does some of the best reviews in the market because you guys use lots of games in your tests, not just one or two games. How hard is to lower game detail settings if you are not getting good playable frame rates? Besides why 4x FSAA? Why not 2X?

I know you will not reply to this but at least take it into consideration in future reviews!

1. If you lower settings in Call of Juarez DX10 benchmark, you will get 14fps, 3fps higher than previously. You won't be able to play, even without FSAA.

2. Your friends had low-quality PSUs/a lot of overclocked components.
Far Cry is the only game that has been tested since 2004.

3. When we are covering 3D performance of a graphics card, we are talking about 3D performance. Offloading video processing is a disputable benefit. Obviously, it is a huge benefit for those with single-core chips that cannot process fullHD VC-1 streams at all. We are critical in regards the lack of proper HD video postprocessing on high-performance GPUs.

4. Virtually all the games coming out in the next few years will have SM3 rendering path, as it is needed for game consoles. Those boards need quality PSUs - that's what we recommend to have all the time.

5. Benchmarking in 0x FSAA, 2x FSAA, 4x FSAA with two ingame quality settings will prolong the time of testing by roughly six times compared to now. We tried to play with settings in DX10 games review, but this did not help cows to fly and current performance-mainstream GPUs to provide decent performance.

1. If that’s the case then like I said blame the developer of the game for laziness. While it’s no problem for the something like 3dmark to run in slide show mode it’s unacceptable in games. The have to do the games to work on today cards not with cards that will be released in 2 years.

2. Yes they are lower quality but they are also cheap. The true is you have to pay a premium price if you want to use those cards, and with the 2600/8600 you don’t.

5. Of course it takes time. But the point is you didn’t review the PowerColor HD 2600 Pro 512MB DDR2 Graphics Card. You did one review of FSAA speed with the card. If your point is to test FSAA performance than do it in a spare article or at least do it right by enabling Transparent AA.

1. You just have to own a proper PSU. It does not mean it should be expensive. Read our PSU reviews: http://www.xbitlabs.com/articles/other/. At the end, your friends had to get two PSUs instead of one and it was more expensive to them than to get a proper one.

5. So, if you cannot play a game, the developer is lazy. But when you see performance numbers that you don't like - the reviewer is lazy.

Take a look at the following test results without FSAA:

- Call of Juarez
- Call of Juarez DX10
- Far Cry with HDR
- S.T.A.L.K.E.R.
- Splinter Cell: Double Agent
- Gothic 3
- Oblivion
- Company of Heroes DX10

They all cannot be played on the Radeon HD 2600 Pro. What a bad world? :)

“They all cannot be played on the Radeon HD 2600 Pro. What a bad world? :)”

But they can Anton! That’s my all point!
You just have to change the setting for that game.
You can use the same setting for a 50$ card as for a 400$ card.
I understand that you want to create a comparing point. And where some doesn’t read the review and just look at the charts and say: hey the 2600Pro is as fast as the nvidia 8800 ultra!

And by detail settings I mean:
-Don’t enable features that make frame rate dunk more (AA).
-Lower in game detail setting options.
-Lower the resolution.

Anton don’t get my words as being bashing your reviews, because your site reviews are the best that can be found on the all web!

For example the last anandtech review of Half life 2 Episode 2 they reviewed the game with 2560x1600! A 30” monitor than almost nobody as one for sure!
Here again the common is?
Exactly what your website reviews 1280x1024 and 1600x1200.

17/19 - 1280 x 1024
20 - 1600 x 1200
Wide 17/19/20 - 1440 x 900
Wide 20/21/22 - 1680 x 1050
Wide 24 - 1920x1200
Wide 30 - 2560 x 1600

You CANT use the same setting for a 50$ card as for a 400$ card.

CANT not can

We do not usually decrease image quality settings in games as different settings affect performance of various GPUs differently. Hence, for correct "apples to apples" comparison we have to keep the highest possible settings all the time.

I've just brought examples of games that could not enable FSAA and were still unplayable on the ATI Radeon HD 2600 Pro. What you say now is to make those games look like DX8-generation titles in order to get the right performance on the DX10 hardware. Why not just stick to DX8 hardware? Because of HD post-processing? Well, I don't think a person with no proper PSU can afford a Blu-ray/HD DVD drive, thus, he/she does not need HD post-processing, whereas those, who build HTPCs will hardly play 3D video games on such machines.

All we are trying to say is that if you want a good graphics card for games, wait for new GPUs.

1280x1024 => all the affordable/installed 17"/19" LCDs.
1600x1200 => advanced LCDs/installed CRTs.
1920x1200 => 24"/27" LCDs (BTW, 24" LCDs start at $500).
2560x1600 => 30" LCDs of enthusiasts.

We can reconsider some resolutions, but we will not get back to 1024x768. It's 2007 and even notebooks support 720p.

What you call enthusiasts I call them very rich people :)

You don’t have any idea how many people still have 17" and 19" CRT monitors to play games, and have high end CPU/GPU with those CRT monitors, some of the guys that replaced them for LCD made just that decision based on available desk space.

Just do some Question Poll and you will see.


“All we are trying to say is that if you want a good graphics card for games, wait for new GPUs.”

Let’s hope November will be the month where everyone will finally surprise us with great products.

Please see separate review covering media playback:

http://www.xbitlabs.com/articles/video/display/video-playback 2.html