XFX 9800 GTX Black Edition 512mb
Synthetic Benchmarks - 3DMark
Published: 13th May 2008 | Source: XFX | Price: £246 |
Futuremark Benchmarks
All of the cards were run through several 3DMark benchmarks. Note that both the 9800 GX2 and the 3 Way SLI setup were both run through an extra benchmark at 1920 x 1200 with 8 x AA.
FutureMark - 3DMark03
3DMark03 is a benchmark that uses mainly a DirectX 8 feature set including several pixel fill-rate tests. I included this test in the review to see an approximation of how well old games will play on the current gen cards, as well as how well each card copes with the fill-rate tests.
All of the cards were run through several 3DMark benchmarks. Note that both the 9800 GX2 and the 3 Way SLI setup were both run through an extra benchmark at 1920 x 1200 with 8 x AA.
FutureMark - 3DMark03
3DMark03 is a benchmark that uses mainly a DirectX 8 feature set including several pixel fill-rate tests. I included this test in the review to see an approximation of how well old games will play on the current gen cards, as well as how well each card copes with the fill-rate tests.
The Black Edition gives a good show in 3dMark03 compared to the stock 9800 GTX and does indeed beat out all of it's competitors by a fair bit.
FutureMark 3DMark05
3DMark05 is a benchmark based on DirectX 9 with more advanced shading and bump-mapping techniques, as well as a tough CPU test integrated.
3DMark05 is a benchmark based on DirectX 9 with more advanced shading and bump-mapping techniques, as well as a tough CPU test integrated.
Again you will notice that the XFX Black Edition 9800 GTX plays out a nice tune, just about crowning over 22k in 05.
FutureMark - 3DMark06
3DMark06 is a more complex 3D benchmark, using many of the more advanced techniques found in DirectX 9.0c such as utilising Shader Model 3.0 and HDR lighting to create a tough benchmark that stresses the GPU and CPU.
3DMark06 is a more complex 3D benchmark, using many of the more advanced techniques found in DirectX 9.0c such as utilising Shader Model 3.0 and HDR lighting to create a tough benchmark that stresses the GPU and CPU.
A common theme throughout the benchmarks, the Black Edition GTX beats out the single GPU cards in our tests. High res benchmarks also show that the card holds good numbers throughout the testing, even when pitted against some more serious scaling.
Most Recent Comments
good review
and nice card
summet that the people with the money, and wanting performance must have then...
and nice card
summet that the people with the money, and wanting performance must have then...
And the people wanting to have overclocked, hand picked GPU's backed up by an excellent warranty and support package
Quote:
|
Originally Posted by name='Kempez'
And the people wanting to have overclocked, hand picked GPU's backed up by an excellent warranty and support package
|
It`s a great review, and I have to be honest with ya`ll, my experiences of oc`ing the 8800GT alpha weren`t the best - and I don`t really put it down totally to the hardware, more the tools required to do it also.
I`d probably be more inclined, in the current climate (things can change of course) to get something above the reference clock and see what minimal % I can go higher than that.
Looking at the price, I don`t think it`s too bad. I can remember my card be around £225 on release.
Good stuff. Good single card imo.
Nice in depth review Kempez. Strange results in FEAR Max fps lol
.
Cue Hudson with his Alien motion tracker : 'It's reading right man look'...
Hicks : 'Well you're not reading it right!'
.Cue Hudson with his Alien motion tracker : 'It's reading right man look'...
Hicks : 'Well you're not reading it right!'
The COD4 results seem strange to me i get basicly the same if not better results on my 320mb 8800 GTS!
Alot can depend on the config settings, not just the res and AA/AF which affects fps..
Agreed, but all cards were tested using the exact same hardware and tested three times so it's hard to argue with what I got. I did also re-test several games again to ensure accuracy
Sometimes you just get one or two results that seem a little odd, but I think generally the results show how fast the card is
Sometimes you just get one or two results that seem a little odd, but I think generally the results show how fast the card is
What weirds me out is when u goto the nVidia driver download thing and there`s a new one - and top of the list of fixes will be something like "9800 cards will now work better with Crysis" - wtf ?!?!
It's not uncommon for developers and manufacturers to work together to improve game performance, that's what Nvidia pay all their coders for 
Quote:
|
Originally Posted by name='webbo'
Alot can depend on the config settings, not just the res and AA/AF which affects fps..
|
My point still stands that the card does not seem to perform very well from what i can see when compared to the lowest 88xx series card.
I think you'll find that the cards held back a little in CoD 4 by the fact that the engine's not exactly a toughie. Gameplay was smooth and silky throughout, but then it will be on all 88XX and 98XX series
Quote:
|
Originally Posted by name='Kempez'
I think you'll find that the cards held back a little in CoD 4 by the fact that the engine's not exactly a toughie. Gameplay was smooth and silky throughout, but then it will be on all 88XX and 98XX series
|
Maybe this is as im hunting for the ability to get 125fps constant at 1920x1200 with AA which atm just aint possible.
Well you see an improvement in the graph from my own benchmarks over the 8800's, just not much
The engine is good which is why so many cards can cope, but no your not going to get that level of FPS, although I would possibly attribute that to CPU/Memory as well
The engine is good which is why so many cards can cope, but no your not going to get that level of FPS, although I would possibly attribute that to CPU/Memory as well
Quote:
|
Originally Posted by name='Kempez'
I would possibly attribute that to CPU/Memory as well
|
The COD/Quake III Engine has a bug in it that lets you jump further depending on FPS you see and 125fps is the sweet spot.
I'll just keep running at 800x600@4AA on my 24" TFT :S dont actually look that bad!
Strange - in the 780a chipset review I did cod4 managed to run on the onboard 8400gs/8500 chip @ 1024x768 with 40fps+ albeit without AA. Testiment to Kempez's view that the engine allows most GPU's to run it ok.I'm surprised you're finding it hard to get a decent fps with the GTS.
76fps and 125fps are as you say the 'sweet spots' so with your screen you would perhaps be better upping the res and lowering the AA and setting the max fps to 76. I have found that on my 3007WFP-HC the AA is barely noticeble at 2560x1600 and certainly not noticble enough that it would warrant me dropping the res.
76fps and 125fps are as you say the 'sweet spots' so with your screen you would perhaps be better upping the res and lowering the AA and setting the max fps to 76. I have found that on my 3007WFP-HC the AA is barely noticeble at 2560x1600 and certainly not noticble enough that it would warrant me dropping the res.
I have a 8800GTS and was running at 125fps more of less constant at 1920x1200 without AA and with some basic tweaks to remove un-needed rubble.
When playing at a high level the slightest FPS drop does become annoying so although i limit at 125FPS really im after 250FPS constant to make sure theres not drops at all.
My system can more than handly it, just inital was stating that im shocked at the lack of improvment from my low end 8800 to this 9800
When playing at a high level the slightest FPS drop does become annoying so although i limit at 125FPS really im after 250FPS constant to make sure theres not drops at all.
My system can more than handly it, just inital was stating that im shocked at the lack of improvment from my low end 8800 to this 9800
CSS Valid
http://www.overclock3d.net/gfx/artic...132909117s.jpg
Link