Do Radeons handle physx better than Nvidia for Borderlands 2? Sounds fishy...

B

biggles

2[H]4U
Joined
Jul 25, 2005
Messages
2,215
http://www.techspot.com/review/577-borderlands-2-performance/page7.html

Quote from conclusion of the article:
"...based on our testing, Radeon cards seem to handle PhysX slightly better than their GeForce counterparts. When running Borderlands 2 at 1920x1200, the HD 7970 only took a 15% performance hit after enabling PhysX (dropping from 72fps to 61fps), whereas the GTX 680 fell 19% from 74fps to 60fps."

That does not sound right. Can anybody give some real world feedback on this?

If true, would be great for Radeon owners like me who were going to pass on the game. If the physx works well on Radeons then I would pick up this game. Also, I am pretty sure other physx games (Batman, Cryostasis, Mafia 2) could not use physx at all with Radeons.

Fyi, this is NOT hybrid physx where there is an Nvidia card installed. The techspot article discusses a game tweak that pushes physx processing to a cpu core. The gist of the article is that multi-core cpu's are now powerful enough to handle physx without a crushing performance penalty.
 
My question would be one of quality. Are the visuals the same? It is possible when it runs on the CPU it automatically reduces the quality.
 
So perhaps Radeon "high physx" is not actually the same graphics quality as Nvidia "high physx" setting?
 
They used a 3960x cpu, since physx is offloaded to the cpu with radeons, I'm sure non-e I5/7's will see a noticable fps hit vs their 6core results.
 
Mozex said:
If physics really are being pushed on to a CPU then wouldn't this indicate that Radeons are just better at rendering Borderlands 2 than GeForce? A similar test would be to just disable physx and compare the frame rate.

Does a 6870 typically perform about the same as a gtx 460 @ 1680x1050? I think the results on this page are physx off.
http://www.techspot.com/review/577-borderlands-2-performance/page2.html
Click to expand...

IIRC, 6870 was generally faster than the 460 and 6850 slightly slower.
 
Using PhysX at max settings in Alice: Madness Returns with a Core i7 920 @ 4 ghz and 6990 in 5760x1080 eyefinity rendered it unplayable once there started being a bunch of debris scattered around. On the other hand, GTX 580 SLI at the same resolution had no problems.
 
Nvidia may have better quality, but that doesn't change the fact that enabling/enhancing PhysX on an Nvidia card is causing way more issues than on AMD cards at the moment. Kind of odd.
 
NVIDIA's PhysX engine doesn't exclude AMD cards from playing the actual game. There are a lot of games out there (e.g. every Unreal 3 engine game) that use PhysX for physics. Some games are developed further with NVIDIA to make effects run on their GPU's only, using hardware based physics (as opposed to software-based through the CPU). Still, there are games like Borderlands 2 where you can enable more PhysX effects regardless if you have an NVIDIA GPU or not.

In my testing, PhysX on "High" runs fine in this game 95% of the time. However, there are a few large battles/fights every so often where the amount of debris will build up and just crush the CPU, and FPS will plummet after awhile from 60FPS into the 20's or lower. This highlights how the parallel processing of a GPU excels at doing NVIDIA's PhysX (big surprise). I haven't found a way to decrease stay time of particles/debris (thereby lightening the load on the CPU), but I haven't played around with it much tbh. I think the PhysX effects look like crap and more unrealistic than base effects offered by having PhysX on "Low."
 
You must log in or register to reply here.
Back
Top