Here are some Half Life 2 numbers for you to look at; they were provided by a reliable source, but I could not verify anything myself so take them with a grain of salt. ATI was running in their DX9 codepath and the mixed mode codepath was used for NVIDIA. No AA/AF was enabled and we're looking at 1024x768 scores:
|
Half
Life 2 Demo
|
Radeon
9800XT
|
NVIDIA
GeForce FX 5950 Ultra
|
| e3_bugbait |
71.4
|
73.9
|
| e3_c17_01 |
57.6
|
57.5
|
| e3_c17_02 |
49.9
|
49.3
|
| e3_phystown |
74.5
|
77.2
|
| e3_seafloor |
53.9
|
53.3
|
| e3_techdemo_5 |
83.5
|
64.5
|
| e3_techdemo_6 |
76.9
|
71.7
|
| e3_under_02 |
77.3
|
71.1
|
If those numbers hold true then things definitely look better than from Half Life 2 day, but we'll reserve judgement until we get the benchmark in house. I just thought you'd like to see what we're seeing, I wouldn't draw any conclusions based on this data yet, just wanted to share :)
Tomorrow is my 8AM day again, and maybe I'll get those tests back that I took on Monday (hopefully not...I'd like tomorrow to be a good day :)...). I'm off to sleep, have a good night everyone. Derek and I will be back to work on Part II tomorrow; I'll update you as soon as I can.
Facebook
Twitter
RSS
102 Comments
View All Comments
Morten - Thursday, October 2, 2003 - link
To #5, if you don't like it, don't read it. Go somewhere else. It's not like these benchmarks are official. Anand even said to take it with a grain of salt. If anything, those numbers are interesting. Guess we'll find out how it stands when the benchmark is released and Anand does a real benchmark.Wotan81 - Thursday, October 2, 2003 - link
Yet an other bougt by nvidia benchmark! LOLAnonymous - Thursday, October 2, 2003 - link
Maybe it's worth waiting until Anand can do some in house benchies. If he doesn't have the demo/time/etc, then it's not his fault he has to use someone else's numbers and not give us much info about them.Anonymous - Thursday, October 2, 2003 - link
Wow I feel sorry for Valve. They wasted 5x the time optimizing the NV codepath compared to the standard DX9 path. All of this time, work, and effort was in vain.In the end, NVidia just optimizes Half Life 2 its own way.
Even though performance is very close, its comparing apples to oranges. You are comparing the NV path, with mixed precision, and missing certain high quality effects such as HDR(which looks AMAZING on screenshots), and are comparing it to ATI's full blown DX9.
Anonymous - Thursday, October 2, 2003 - link
You might as well just benchmark Nvidia cards at 640x480 and compare them to ATI cards at 1200x1600. Who is this source and why didn't they benchmark BOTH the FX and 9800 at full precision?Sazar - Thursday, October 2, 2003 - link
it is using pp codepath... as anand says.. mixed precision... if it is like the nv35 therefore it will be using fp16/32 likely...I do however find it quite shocking that there is no additional column for dx9 codepath performance figures by anand for the nv38... that comparison and the subsequent performance gains would be what we would be looking @ for comparison...
Anonymous - Thursday, October 2, 2003 - link
Hopefully valve will keep those release updates along with HL2 that stop cheating as they find them when time goes on. Very very odd numbers if you ask me, they don't jump that high "automagicly" and we all know nvidia's hardware is not up to it. like every one is saying, there's going to be some serious IQ problems.Anonymous - Thursday, October 2, 2003 - link
Performance like this doesn't come from thin air. Probably reducing everything to as low as FX12. Oh and you also failed to mention that these scores are without certain features like HDR enabled, since FX cards can't do it at all. Maybe throw in some static clip planes specifically for the benchmark camera sequences.....Its unbelievable that you would post these numbers coming from a "reliable source" without any sort of IQ comparison from tests in lab. And what source might that be....Derek Perez? Brian Burke? Thats Anand for you
Anonymous - Thursday, October 2, 2003 - link
i will look fine , maybe not as nice as ati's but we dont know that yet , plus ive seen some pics of ps 1.4 and ps 2.0 and to tell you the truth i could barely tell the diff {tomb raider}, it was there but barely notisableAnonymous - Thursday, October 2, 2003 - link
I am tired of these BS.Why these major hardware sites are too ready to believe whatever nvidia spews and spread false around the mass?