Finally, we wanted to compare the new drivers Nvidia has made available for the release of 3DMark Vantage - ForceWare 175.12 beta - to the older ones with the GeForce 8800 Ultra.
The differences are significant. While the GPU score is "only" 8% higher, New Calico, and especially Feature Tests 3 (the one that uses parallax mapping, and on which the Radeon cards literally collapse) and 6 (Perlin noise) were really off the chart, without any possibility of doubt as to the specific nature of the optimizations. We should remind you here that, according to the Futuremark charter that validates the drivers for 3DMark, a driver must not have been optimized specifically for the benchmarking application.
AMD made their drivers for Futuremark (Catalyst Hotfix 8.471.1) available too late for us to use them in the preceding tests, but we were able to compare them to the previous Catalyst 8.4 WHQL version with a Radeon HD 3870 X2:
This time, the gain was even more impressive, both with the overall score and especially with Feature Test 3, which again was off the charts, with the Radeon HD 3870 X2 reaching 15.93 fps and taking the lead over all the other cards, despite its earlier abysmal performance in this test! The gains for Feature Tests 4 and 5 are also much too great to be legitimate. And indeed the rendering looked slightly corrupted to us, especially with New Calico, with horizontal lines appearing.
Current page: Optimizations?Prev Page Overall Scores and Optimization Next Page No More Free Versions!
Stay on the Cutting Edge
Join the experts who read Tom's Hardware for the inside track on enthusiast PC tech news — and have for over 25 years. We'll send breaking news and in-depth reviews of CPUs, GPUs, AI, maker hardware and more straight to your inbox.
Why should I pay good money tobe shown that my close to state of the art overpriced hardware sucks.Reply
Because chicks love it when you post the high numbers from your overclocked machine in forums.Reply
I think they might have underestimated how cheap a lot of the PC community is. $20 for a utility to get some arbitrary number seems a bit high to me.
What an incredibly stupid move. In the long run futuremark will make less money because, as Florian says, popular support for this product is indispensable. Now that Tom's is shooting it down, it has already lost a significant amount of share. This product has been all about hype, and there seems to be little hype for it this time around. I'm certainly not paying for what appears to be more of the same. With two years instead of just one, and DirectX 10 instead of DirectX 9, I would have expected something better.Reply
When I tried to get my results I get this fatal error: http://img508.imageshack.us/img508/6963/captureir4.pngReply
Well this is the first and only pure DX10 title you can test on. Every other (real games) use DX9 engine with DX10 makeup. (Ofcourse, because it would be economical suisade to make only DX10 game at this moment (too few Vistas out there))Reply
So is you want to know if you graphic card is any good in DX10 environment, this is the only test you can use. What it's good for... Well not much at this moment, because it will take years untill we will see pure DX10 game engines, because the DX9 support is too important to the game makers!
But one day those DX10 games will come out, so until then this is the one and only test that you can use to test yours gpu's DX10 power...
duuuuuuuuuur huuuuuuuuur lets use sintetic benchamrks instead of real world game testsReply
seriously i am happy tomshardware wont be using 3dmarks newest applet anymore =D
Its amazing what some Hotfix drivers can do. 477% on a useless test is still pretty impressive. Imagine what ATI and Nvidia could do with their present day lineup if they took the time to optimize the drivers for each game. I mean REALLY optimize the drivers. Nvidia should be much better at this since they've had so long to optimize for the G80. ATI should probably spend more time optimizing than trying to get a new driver out each month. Their drivers sometimes degrade performance.Reply
As far as 3DMark Vantage goes, its useless in the real world. We should have Unreal 3 powered benchmarks since so many games actually use that game engine.
If 3Dmark Vantage is anything like PCMark Vantage, stay away. It's the most aggravating, stupidest program I've ever used. I wasted $20 on this piece of crap. To view your results you are directed to a web site that has literaly hundreds of pop-up adds all mixed in with your results. Realy, realy stupidReply
I can run free TRIAL VERSION asm many as I want - just click UPGRADE LATER and RUN !Reply
with fire GL v. it won't work, although most OTHER test do work, not vantage.Reply
Signed:PHYSICIAN THOMAS STEWART VON DRASHEK M.D.