• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

PhysX FluidMark results thread

13,735.....This is my first run, no control panel or desktop tweaks, straight in! funny thing is, if I increase core/shaders higher, my score gets lower :confused:

I am guessing here the CPU speed makes no difference at all as I only had mine at 3.2gig on this run as I am testing a passive cooler on my CPU to see how it cools.
 

Attachments

  • untitled.JPG
    untitled.JPG
    165.5 KB · Views: 419
Last edited:
Very nice!
 
Last edited:
q6600@3.33 ghz
8800 gts (g80) @ 681/1006
9517
 

Attachments

  • physxmark.JPG
    physxmark.JPG
    113.3 KB · Views: 336
I'm only getting around 8000 with dual 280GTXs. Is this because I'm running 64bit Vista?
 
I'm only getting around 8000 with dual 280GTXs. Is this because I'm running 64bit Vista?

Try it with just one, bet the score is higher.
 
not much of a difference
 
Need MOAR megahurtz !!
9648
 

Attachments

  • physx.JPG
    physx.JPG
    118.7 KB · Views: 388
I'm only getting around 8000 with dual 280GTXs. Is this because I'm running 64bit Vista?

Vista has some effect, average -3K compared to XP, but beyond that it seems this bench is somewhat CPU dependant too...best I've gotten with an OC'd GTX260 and Q6600 at 3.6 is 8.3k in Vista x64...I think some of it is the new PhysX support and drivers. That and maybe the program is Beta and not intended for x64 yet...things can change in upcoming versions...just look at their Furbench...the differences in performance for some cards between version 1.0 and 1.4 can get pretty drastic....so don't worry too much about it.

:toast:
 
1st run was 2x 8800gs with pci-e @ x8. 2nd run was 1x 8800GTS @ x16 pci-e. No driver changes between runs. It's not a fluke, as i made the run three times and posted my best run. All runs were within 100pts of each other. The GPU has not topped out yet, but it is in another case folding right now. In the next week I may give it another run. We will see what it can do....
lol the 13054 run was at x8 ;)
 
yeah it does seem that this is limited in vista - we'll give them some time to work the kinks out, but its still craploads faster than purely CPU.
 
ahhhhhh I see, it's a 32bit only app........that solves it Energy!
 
ahhhhhh I see, it's a 32bit only app........that solves it Energy!

That's a good point. I though originally I couldn't install it b/c it said 32-bit, but after I did and it worked I forgot all about it. I guess that's coming into play too. So operating system is actually currently the largest determining factor. :wtf:

Oh and cpu score does play a small role, I got 300 extra points (in vista x64 no less) for a 400mhz oc on a conroe.
 
Yes, give me the criteria you want the test runs completed with.
Keep in mind it may not be until later this week...

hmmm criteria, graphicscard max stable oc, or like the settings in your bets run, different cpu speeds, like stock clock, mild oc and max oc AND if you have time some comparisons between different fps speeds but with the "same" cpu clock ... it would be really nice if you had time and if you feel like it :rockout:
 
Slightly better, almost 14000 now.
 

Attachments

  • fluidmark.JPG
    fluidmark.JPG
    163.3 KB · Views: 409
sorry about not updating. I have been moving and don't have my computer working right now. it won't post. will update when i am on my main rig don't like this one that well
 
Broke 16000 with a single card, done a few tweaks in my BIOS and came up with this, it's amazing how these cards accelerate as you add on the shader clocks..............
 

Attachments

  • fluid.JPG
    fluid.JPG
    181.8 KB · Views: 396
Broke 16000 with a single card, done a few tweaks in my BIOS and came up with this, it's amazing how these cards accelerate as you add on the shader clocks..............

Wow, great oc on that tatty. What kind of bois mod's are you talking about, bios vmod? And is that on stock cooling?
 
Wow, great oc on that tatty. What kind of bois mod's are you talking about, bios vmod? And is that on stock cooling?

Yes stock cooling, she runs fine although to bench I have to up the fan to 100% but game at 60-70% depending on game......I tried the BIOS 1.18V Mod and dint think it was working so flashed back to stock, then a friend of mine who has a 260 said that he read on a forum, where a member basically ran his GPU underspeed at 550mhz core for 3 days, then did the 1.18V mod, so I tried it and it seems to have let me have a bit more on the core but more importantly for this bench....the shaders.....it's a set of shaders that do the PhsicX so the really important thing in this bench seeems to be high shader speed.......(probably why the G92 GTS's are doing well)I think I just got lucky with the overclock, it's not 100% stable though, I can play Source and COD4 (for 20 minutes) but not do a 2006 run, need to drop to 785 with shaders linked for that.
 
Yes stock cooling, she runs fine although to bench I have to up the fan to 100% but game at 60-70% depending on game......I tried the BIOS 1.18V Mod and dint think it was working so flashed back to stock, then a friend of mine who has a 260 said that he read on a forum, where a member basically ran his GPU underspeed at 550mhz core for 3 days, then did the 1.18V mod, so I tried it and it seems to have let me have a bit more on the core but more importantly for this bench....the shaders.....it's a set of shaders that do the PhsicX so the really important thing in this bench seeems to be high shader speed.......(probably why the G92 GTS's are doing well)I think I just got lucky with the overclock, it's not 100% stable though, I can play Source and COD4 (for 20 minutes) but not do a 2006 run, need to drop to 785 with shaders linked for that.

Interesting. You keep saying you get lucky but I think your being modest. :laugh: You seem to always end up w/ some sick oc's that no-one else around here can match w/ the same means (air and stock cooling). I guess little tricks like that one you picked up there will do it though.

I think shaders do play a big role in physX, but I'm not convinced that's the biggest factor w/ this bench, unless it's just crippled by x64 and vista, I have seen very minimal gains from shader oc w/ my set-up.
 
just so you guys know i am not going to be updating this list anymore now that i am back in school and i am going to be picking up a job soon.(don't ask where i am shamed)


IF ANYONE WANTS TO TAKE OVER THIS THREAD PM ME AND I WILL GET A MOD TO HELP US OUT.
Thanks and sorry for caps
 
just so you guys know i am not going to be updating this list anymore now that i am back in school and i am going to be picking up a job soon.(don't ask where i am shamed)


IF ANYONE WANTS TO TAKE OVER THIS THREAD PM ME AND I WILL GET A MOD TO HELP US OUT.
Thanks and sorry for caps

YGPM
 
tatty i sent a PM to dan to see if he can do like thermo did in the teen thread if he can then you will have the 2nd post of the thread to update it.
 
By my score its easy to see that this was designed around Nvidia cards and drivers...

Intel e7200 @ 3.6 ghz
Visiontek 3870 @ 931/1395
Asus P1 PhysX card


2jevhic.jpg
 
By my score its easy to see that this was designed around Nvidia cards and drivers...

Yep! It was designed using NV's GeFroce PhysX Engine!

PhysX FluidMark is a physics benchmark based on NVIDIA PhysX engine. This benchmark performs a fluid simulation by imitating the renderering of lava. Real physics parameters such as viscosity are used. SPH (Smoothed Particle Hydrodynamics) algorithm is enabled to increase the realism of the simulation.

Source: http://www.ozone3d.net/benchmarks/physx-fluidmark/

But if sources are true that NV and ATI are working together on physx for both sides...then ATI guys will eventually benefit more and more from what is happening now with NV cards.

:toast:
 
Back
Top