http://www.rage3d.com/r3dtweak/
come on you guys,have a try to see how your 9700 performances without PS 1.4
come on you guys,have a try to see how your 9700 performances without PS 1.4
the new rage3d tweak can force shader version in DX
- Thread starter DOOM III
- Start date
From tests I did earlier today.
9700 at 1.1, 1.4, and 2.0
Multicompare link
I tried to test my 8500 at 1.1, but at game test 3 it hangs up every time with the Catalyst 3.1 drivers at least. Haven't had time to figure out if I can get around it yet.
Edit: fixed link
9700 at 1.1, 1.4, and 2.0
Multicompare link
I tried to test my 8500 at 1.1, but at game test 3 it hangs up every time with the Catalyst 3.1 drivers at least. Haven't had time to figure out if I can get around it yet.
Edit: fixed link
Sorry to take this off topic. Ichneumon, I saw that you were able to run the 3rd sound test and are using an audigy. Are you using an audigy 2? I'm using an audigy 1 and get the message that it is not supported. Has anyone been able to get the 3rd sound test to run using an audigy1?
THe_KELRaTH
Regular
It's interesting to see that comparing the 9700 using PS 1.1 v GFFX Ultra with non modded drivers the GFFX still falls along way behind.
24.9 v 19.4
22.8 v 16.8
And thats ignoring the CPU difference.
24.9 v 19.4
22.8 v 16.8
And thats ignoring the CPU difference.
DOOM III said:http://www.rage3d.com/r3dtweak/
come on you guys,have a try to see how your 9700 performances without PS 1.4
Can I force PS 2.0 in Morrowind?
I'm not sure if this was your intent, but I'm going to pick on you anyways:
There seem to be a lot of people (or maybe it's just Kyle Bennett over and over) discounting the new GFfx drivers, seemingly for the sole reason that they provide a really big boost to 3DMark03 performance.
If it turns out the performance gains show up only in 3DMark03 (and we have to be sure to compare to other shader-heavy code before making a judgement), or if some IQ anomaly or something along those lines is discovered, then we have plenty of time to start yelling. If--as is seemingly likely--they switch the "default" for PS 2.0 from FP32 to FP16, then apparently this will have to be changed, although with a misprint in the spec we shouldn't blame Nvidia too harshly for this, especially if they fix it in timely fashion.
The point is: these are still pre-release drivers. Nvidia should still be getting huge performance gains out of them (especially considering where they started from), particularly in the PS 1.4-2.0 functionality which is of course entirely brand new for NV30. There is no reason (yet) to assume these drivers use anything other than legit, generally applicable (at least to shaders) optimizations.
(More to the point: even a glance at the "before" and "after" score reports at [H] indicates the improvements are completely confined to PS 1.4 and 2.0, with the vast majority coming in the former. This is entirely to be expected--particularly given how poor the performance was to start--and chances are very, very good that all the optimizations apply equally well to PS 1.4 shaders other than those used in 3DMark03.)
In any case, there's certainly no reason to pretend that the previous drivers--which put the GFfx Ultra at Radeon 9500 NP speeds in games 2, 3 and 4--are in any way more "correct" or more respresentative of "normal" GFfx performance.
THe_KELRaTH
Regular
While I appreciate I shouldn't jump to any firm conclusions the one thing that stood out between the 2 drivers was that the PS2.0 test only increased by 3.5fps and the vertex shader score increased by 4.5fps (9.3 to 13.8fps) at 1024x768 but Mother Nature test improved quite significantly - 18.7 to 30.7fps.
GT4 uses PS 1.4 shaders (for the ground) in addition to PS 2.0 (for water and sky). So one quick hypothesis is that the new drivers improve PS 2.0 performance a bit and PS 1.4 performance a lot.
Also there are obviously VS 1.1 improvements (as per vertex shader test), which according to Nvidia's Whine Paper would have added to the large improvements in GT2 and 3 anyways.
Also there are obviously VS 1.1 improvements (as per vertex shader test), which according to Nvidia's Whine Paper would have added to the large improvements in GT2 and 3 anyways.
BRiT said:jjayb said:
I'm using the built-in audio provided by Nvidia's SoundStorm on a nForce2 motherboard and I'm able to run all 3 tests. I'm glad I ditched my Audigy-1 card.
Heh, not to continue this OT in this thread... but I thought the soundstorm audio on my Asus A7n8x Deluxe sounded like arse. So I got the Audigy2.
Similar threads
