Physics cards

Forum Graphic & Displays : Graphics Cards - Physics cards

Tom's Hardware: Over 1.4 million members in 6 different countries available to answer all your high-tech questions. Sign up now! Its free!
Word :    Username :           
 

Hi guys I have a small question.


I'm have trouble finding some benchmarks about dedicated physics cards on pci-ex 1x, x4, x8, x16.Do you guys think there would be a difference in any of these?

Sponsored Links
Register or log in to remove.

Perhaps 1x would be limiting, but I'd assume 4x and above (especially if PCI-E 2.0) would not cause issue. What sort of Physics cards are you referring to? PhysX cards, or something else? If the former, just consider it a low end GPU.


Message edited by EXT64 on 11-10-2009 at 04:45:31 AM
------------------------------ New Build (Under Contruction): Intel Core i5 750 > Gigabyte GA-P55-UD4P > 2x2GB GSkill 1600MHz CL7 1.65V > Samsung Spinpoint F3 1TB > Antec 300 Illusion > Asus 4850 512MB w/AM Cooler > Corsair 650HX > Hyper 212 Plus > See other Builds in Member Config
Reply to EXT64

Well I saw some comments on google where people running on high resolutions like 1920x1200 and above with all eye candy.Some said that they use 8800,9800 cards on the x16 so that they wouldn't see bottleneck on the card.So thats why I was wondering has there been a benchmarks on that topic on any resolutions.

Reply to jasonx

Well, here is a site showing a comparo of CPU PhysX, PhysX on the rendering card, and PhysX on a dedicated card from a 8400 to a 9800. From the looks of it, not much is needed. A 9600 looks like the approx 'sweet spot'. A super high end dedicated doesn't seem to add much, but I just glanced over the results, so you probably can gain more by looking more closely. However, if only a 9600 is needed, I think 4x (maybe) or 8x (or more) is probably all that is necessary.

 

Edit: Link: http://www.pcgameshardware.com/aid [...] ws/?page=2


Message edited by EXT64 on 11-10-2009 at 05:00:48 AM
------------------------------ New Build (Under Contruction): Intel Core i5 750 > Gigabyte GA-P55-UD4P > 2x2GB GSkill 1600MHz CL7 1.65V > Samsung Spinpoint F3 1TB > Antec 300 Illusion > Asus 4850 512MB w/AM Cooler > Corsair 650HX > Hyper 212 Plus > See other Builds in Member Config
Reply to EXT64

Thx for the site.It would be cool thought if someone did a test on this just to see if physics card would operate better on pci-e 1.0 or 2.0 and also x1,x4,x8,x16.

Reply to jasonx

Well, for 1.0 vs 2.0, you can just (approximately) double the x## to go from 1 to 2 (since 2 has 2x the bandwidth). As for the 1x, 2x, 4,etc., assuming it follows the graphics card trend (which since they are graphics cards, though they may be less bandwidth limited, so this link may be the worst case) slightly older/not top of the line cards did quite well on PCI-E bandwidths below x16 and even x8

Older, so 1.0:
http://www.tomshardware.com/review [...] ,1572.html

Newer, so 2.0:
http://www.tomshardware.com/review [...] ,1915.html

------------------------------ New Build (Under Contruction): Intel Core i5 750 > Gigabyte GA-P55-UD4P > 2x2GB GSkill 1600MHz CL7 1.65V > Samsung Spinpoint F3 1TB > Antec 300 Illusion > Asus 4850 512MB w/AM Cooler > Corsair 650HX > Hyper 212 Plus > See other Builds in Member Config
Reply to EXT64
Tom's Hardware > Forum > Graphic & Displays > Graphics Cards > Physics cards
Go to:

There are 1114 identified and unidentified users. To see the list of identified users, Click here.

Sponsored links
  • Ask the community now
  • Publish
Ad
They won a badge
Join us in greeting them