• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Project cars benchmarks

http://forums.anandtech.com/showpost.php?p=37389081&postcount=101

He has been asked to do it again with the Weather effects on as that uses Physx as well.

Let me:

Ahem, 32 cars, rain, max settings, CPU vs GPU physX, can't really get any worse:

5a1a7e1d_pcars_testi2_annotated.PNG


3 runs for both CPU and GPU physX to minimize the variance between runs.

And here's the worst case scenario run that I ran 6 times in total:


Right at the start of a race with all the cars on the screen and lots of water effects everywhere because of it. Also the player car visible so you have to render the water effects for that as well including all the rain drops on the car.

(also excuse the keyboard driving)


This would point at it not being a game issue.
 
The first game where Windows 10 makes a huge difference.

Wonder why that is?

Is it a boost overall for everyone or just one vendor? If its one vendor could indicate they been working more on win 10 driver than win7/8
If its both could be the O/s handling things better overall
 
Last edited:
I did that and recorded a session with GPU PhysX, then set the CP up for the CPU to deal with PhysX and ran the same lap with the same conditions. Everything is fully maxed out and at 1440P....


As you can see, there is no real difference between the 2 and by all accounts of what some are saying, performance should be tanking badly with the CPU handling the PhysX.

Quoting myself to backup what Alatar said. You can see there is no difference (or anything boticeable) with PhysX being run on the CPU (how AMD have to run it), so that certainly isn't the cause and the whole scenario points to AMD missing the opportunity to work with the devs. No GameWorks gimping that likes to be the backbone of any AMD users argument. You can see that Windows 10 has it working well and that must point to drivers.

It would be nice for once to see some honesty and people look further than it always being nVidia's fault!
 
You guys using nvidia control panel to set GPU or CPU Physx? Or is this a game setting? If it's not a game setting then selecting GPU will make zero difference because the game will be CPU Physx hard coded.

Has for Windows 10 running it better, maybe windows 10 amd driver is less overhead freeing more CPU usage to boost frame rate. It would be nice to see some more windows 8.1 vs 10 comparisons.
 
You guys using nvidia control panel to set GPU or CPU Physx? Or is this a game setting? If it's not a game setting then selecting GPU will make zero difference because the game will be CPU Physx hard coded.

Has for Windows 10 running it better, maybe windows 10 amd driver is less overhead freeing more CPU usage to boost frame rate. It would be nice to see some more windows 8.1 vs 10 comparisons.

They used the same driver...
 
Point is that if the game was using GPU physX on Nvidia cards and CPU physX on everything else (and causing big performance differences) it'd be extremely easy to test by forcing CPU physX from the NV control panel.

And that's what people have done. And since there's no difference the only logical conclusion is that either the game runs CPU physX for everyone all the time or that using GPU PhysX has almost no performance advantage.

But NV using GPU PhysX and AMD CPU PhysX is clearly not the reason for the performance gap.

And besides, the game has already been conclusively proven to run CPU PhysX only for all manufacturers:

http://forums.anandtech.com/showpost.php?p=37389404&postcount=140

There's no GPU PhysX implementation in the game.
 
Point is that if the game was using GPU physX on Nvidia cards and CPU physX on everything else (and causing big performance differences) it'd be extremely easy to test by forcing CPU physX from the NV control panel.

And that's what people have done. And since there's no difference the only logical conclusion is that either the game runs CPU physX for everyone all the time or that using GPU PhysX has almost no performance advantage.

But NV using GPU PhysX and AMD CPU PhysX is clearly not the reason for the performance gap.

And besides, the game has already been conclusively proven to run CPU PhysX only for all manufacturers:

http://forums.anandtech.com/showpost.php?p=37389404&postcount=140

There's no GPU PhysX implementation in the game.

Am calling that this game is CPU physX hard coded.. Its not the first game to do this. Arma 3 has hard coded CPU physX..
 
No matter what the reason turns out to be what it ultimately boils down to is the way AMD won't entertain developer invites during development, if you don't bother pulling a chair up to the table like Intel and NVidia both did then don't be surprised when the others who do attend shaft you. AMD customers are the ones who are ultimately hurt because AMD just make their usual excuses absolving themselves of all responsibility and their most loyal fans lap it up, so nothing will ever change.

Expect a patch in about 3mths time if you're lucky because 3mths ago AMD were still living in cloud cuckoo land expecting Intel and NVidia to do what's in AMD's best interest.
 
Has anyone tried those modded win10 drivers on win8.1? People are reporting roughly 20% drawcall performance increases on 3dmark api test over 15.4 drivers. Noticeable differrences in cpu limited game-scenarios aswell.
 
Has anyone tried those modded win10 drivers on win8.1? People are reporting roughly 20% drawcall performance increases on 3dmark api test over 15.4 drivers. Noticeable differrences in cpu limited game-scenarios aswell.

I haven't I dont play around with modded drivers but yes I also seen reports nice gains across the board.
 
Back
Top Bottom