180.48 WHQL drivers released

R

RyanH

Limp Gawd
Joined
Feb 22, 2006
Messages
391
So these are the Big Bang II drivers I presume. It says multi-monitor SLI is enabled on these too. Also:

* Boosts performance in numerous 3D applications. The following are some examples of improvements measured with Release 180 WHQL drivers vs. Release 178 WHQL drivers (results will vary depending on your GPU, system configuration, and game settings):
o Up to 10% performance increase in 3DMark Vantage (performance preset)
o Up to 13% performance increase in Assassin's Creed
o Up to 13% performance increase in BioShock
o Up to 15% performance increase in Company of Heroes: Opposing Fronts
o Up to 10% performance increase in Crysis Warhead
o Up to 25% performance increase in Devil May Cry 4
o Up to 38% performance increase in Far Cry 2
o Up to 18% performance increase in Race Driver: GRID
o Up to 80% performance increase in Lost Planet: Colonies
o Up to 18% performance increase in World of Conflict
Click to expand...

Hope to see some of those gains especially warhead.
 
Good, can't wait to load 'em up! My poor GTX260 needs all the help it can get with this new 30 incher.

Are these up on nV's download site?
 
how can there be a 38% increase in farcry2, but only a 10% increase in warhead? don't they use the same engine?
 
darkpaw said:
No they don't http://enthusiast.hardocp.com/article.html?art=MTU3MSwxLCxoZW50aHVzaWFzdA==

Thanks for the link. I hope they fixed the cursor bug I was having in the beta version.
Click to expand...

huh. i actually read that article too. i probably should have paid more attention.

as a side note, seems kind of pointless to spend so much time and money to develop such a powerful yet taxing engine (cryengine2) if you're not going to use it in one of your biggest games. yeah, they have other games planned to use cryengine2, but farcry is what put them on the map.
 
stainremover said:
huh. i actually read that article too. i probably should have paid more attention.

as a side note, seems kind of pointless to spend so much time and money to develop such a powerful yet taxing engine (cryengine2) if you're not going to use it in one of your biggest games. yeah, they have other games planned to use cryengine2, but farcry is what put them on the map.
Click to expand...

Crytek, the developers of Far Cry 1 and Crysis did not develop or was in anyway involved with Far Cry 2. I would guess the actual rights to the "Far Cry" franchise is owned by Ubi Soft and not Crytek but I could be wrong.
 
running 3dmark vantage right now, I will report back my findings when I am done.
 
Ok for my first run with my setup I get:

3dmark score of T22380

cpu score: 47975
gpu score: 19001
 
25% gain in DMC4 is not happening, it actually LOWERED my score
 
most of the gain stuff is BS. you never actually see those gains. And they never say at what setting those gains were.
 
wolf2009 said:
most of the gain stuff is BS. you never actually see those gains. And they never say at what setting those gains were.
Click to expand...

well DMC4 at max AA @1920 by 1200 sucks now on scene two, I need to play it but it never had a 30fps avg before
 
wolf2009 said:
most of the gain stuff is BS. you never actually see those gains. And they never say at what setting those gains were.
Click to expand...
When they say "up to" it means you could potentially see those gains. Those numbers are only an indicator of what is possible depending on your system and game settings.
 
dark_reign said:
When they say "up to" it means you could potentially see those gains. Those numbers are only an indicator of what is possible depending on your system and game settings.
Click to expand...

maybe it really is giving a performance hit, and you get those types of gains when you drop in a gtx280 for physx?
 
Once I get home from work I am putting my old 8800GTS 640MB back into my system and installing the new drivers. Now I just need to find a game that uses it.
 
Would someone mind testing Digital Vibrance to see if it works (the slider will go through all the small percentages, not just 33%, 66% and 100% and the color saturation actually changes)?
 
The "up to 38 percent" thing is probably comparing the 180.48 WHQL drivers to the last WHQL release, 178.24. There were some substantial improvements in the betas in between.

Some people may be unaware that the PhysX is set to render physics on the video card. PhysX can be set to render on your CPU instead of the video card. If your video card is maxed out while your CPU is not (which is usually the case with a high-speed dual-core or quad-core), then rendering PhysX on the video card can actually slow you down.

I noticed when running the Far Cry 2 benchmark test that if I set PhysX to render on the CPU, my FPS improved.
 
evilsofa said:
The "up to 38 percent" thing is probably comparing the 180.48 WHQL drivers to the last WHQL release, 178.24. There were some substantial improvements in the betas in between.

Some people may be unaware that the PhysX is set to render physics on the video card. PhysX can be set to render on your CPU instead of the video card. If your video card is maxed out while your CPU is not (which is usually the case with a high-speed dual-core or quad-core), then rendering PhysX on the video card can actually slow you down.

I noticed when running the Far Cry 2 benchmark test that if I set PhysX to render on the CPU, my FPS improved.
Click to expand...

how to do you change it to render physx on the cpu over gpu?
 
evilsofa said:
The "up to 38 percent" thing is probably comparing the 180.48 WHQL drivers to the last WHQL release, 178.24. There were some substantial improvements in the betas in between.

Some people may be unaware that the PhysX is set to render physics on the video card. PhysX can be set to render on your CPU instead of the video card. If your video card is maxed out while your CPU is not (which is usually the case with a high-speed dual-core or quad-core), then rendering PhysX on the video card can actually slow you down.

I noticed when running the Far Cry 2 benchmark test that if I set PhysX to render on the CPU, my FPS improved.
Click to expand...

that was my first instinct but i was swayed otherwise by this:

http://www.firingsquad.com/hardware/physx_performance_update/

haven't done actual testing myself.

and yes, as stated in the OP it is vs the old WHQL drivers
 
DoomzDayz said:
that was my first instinct but i was swayed otherwise by this:

http://www.firingsquad.com/hardware/physx_performance_update/
Click to expand...

I think you misunderstood the test. They were comparing how fast nVidia GPUs, PhysX PPU cards and CPUs execute the PhysX code. To test how fast the GPU runs the PhysX code, they had to choose situations in which the GPU was not maxed out rendering FPS.

Whether or not the GPU does the PhysX code faster than the CPU is irrelevant when the GPU must do the PhysX code instead of putting out FPS.

Let me put this another way: if the CPU's second core is idle while the GPU is maxed out on FPS, then you will want to do the PhysX code on the CPU's second core rather than on the GPU, because if you run it on the GPU, you do so at the cost of FPS.
 
evilsofa said:
I think you misunderstood the test. They were comparing how fast nVidia GPUs, PhysX PPU cards and CPUs execute the PhysX code. To test how fast the GPU runs the PhysX code, they had to choose situations in which the GPU was not maxed out rendering FPS.

Whether or not the GPU does the PhysX code faster than the CPU is irrelevant when the GPU must do the PhysX code instead of putting out FPS.

Let me put this another way: if the CPU's second core is idle while the GPU is maxed out on FPS, then you will want to do the PhysX code on the CPU's second core rather than on the GPU, because if you run it on the GPU, you do so at the cost of FPS.
Click to expand...

he basicaly means:

buy an 8400GS, and your set ;)
 
Now I'm curious; I see the 8400GS going for as little as $23 on newegg. Has anyone actually plopped an 8400GS in their system just for PhysX, and does it show any performance improvement over just letting the CPU do it?
 
Can I use this 180.48 driver with my 7950 GX2??

I would like to upgrade my driver from its current 178.34 version..

Plz tell me guy, I don't want to end up without any display after installing this new driver..

Thanks.
 
evilsofa said:
Now I'm curious; I see the 8400GS going for as little as $23 on newegg. Has anyone actually plopped an 8400GS in their system just for PhysX, and does it show any performance improvement over just letting the CPU do it?
Click to expand...

i think some people on the evga forums had tried this with an earlier version and showed fps drops. the link was posted earlier in another thread here, around the beginning.
 
evilsofa said:
Some people may be unaware that the PhysX is set to render physics on the video card. PhysX can be set to render on your CPU instead of the video card. If your video card is maxed out while your CPU is not (which is usually the case with a high-speed dual-core or quad-core), then rendering PhysX on the video card can actually slow you down.

I noticed when running the Far Cry 2 benchmark test that if I set PhysX to render on the CPU, my FPS improved.
Click to expand...

Far Cry 2 is not a PhysX game, it does not make use of PhysX .

So it doesn't matter where you set the physX to render. This nullifies your statement .
 
evilsofa said:
I think you misunderstood the test. They were comparing how fast nVidia GPUs, PhysX PPU cards and CPUs execute the PhysX code. To test how fast the GPU runs the PhysX code, they had to choose situations in which the GPU was not maxed out rendering FPS.

Whether or not the GPU does the PhysX code faster than the CPU is irrelevant when the GPU must do the PhysX code instead of putting out FPS.

Let me put this another way: if the CPU's second core is idle while the GPU is maxed out on FPS, then you will want to do the PhysX code on the CPU's second core rather than on the GPU, because if you run it on the GPU, you do so at the cost of FPS.
Click to expand...

this is what i initially thought, however, looking at it again...

software has 13 fps. this means upping the resolution, the fps is 13. with the gpu physx, upping it will start to lower it from 30+ but I seriously doubt it will go below 13 fps unless you go very high. nevertheless, its 13 fps (CPU) or above 13 fps with SOME AA/AF/higher resolution with GPU physx. granted at max settings with gpu physx it could be below 13fps and cpu physx would be better but i wouldn't want to play at that.

now i know ut3 is a quad core optimized game so this may not be the case in other games but there is no data showing that in this article.
 
DoomzDayz said:
this is what i initially thought, however, looking at it again...

software has 13 fps. this means upping the resolution, the fps is 13. with the gpu physx, upping it will start to lower it from 30+ but I seriously doubt it will go below 13 fps unless you go very high. nevertheless, its 13 fps (CPU) or above 13 fps with SOME AA/AF/higher resolution with GPU physx. granted at max settings with gpu physx it could be below 13fps and cpu physx would be better but i wouldn't want to play at that.

now i know ut3 is a quad core optimized game so this may not be the case in other games but there is no data showing that in this article.
Click to expand...

UT3 doesn't count as a game lol.
 
spaceman said:
UT3 doesn't count as a game lol.
Click to expand...

point taken, lol. i've never played it. there are 2 other "games" they used with the same results. are there any other comparisons?

thankfully i was out of the computer hobby for the past couple years and missed the whole physx phenomenon only to find it integrated in my new video card
 
In case it hasn't been mentioned. Lord of the Rings Online players should still avoid these drivers if they use DX10. The shadows are still screwed up.
 
The setup program won't install on my rig. Says I don't have compatible HW. I have a 8600GT with 178 WHQL.
 
You must log in or register to reply here.
Back
Top