NEW Benchmark scene (Max 8 Vray RC3 required)

[arobbert]

HP Z600
2x E5530 2,4Ghz
8 GB RAM
Nvidia QuadroFX 1800
XP Pro 64+
Max2010
V-Ray Sp3a

52,4 sec



sorry, wrong gamma but rendetime is the same...

edit:

at home with my Dell Precision M65 Notebook 2 Ghz Dual Core, 4GB Ram and Nvidia Quadro 350M

5,33 min

How many passes did you set for the light cache? Are you hyper threading?

[stellwagenc]

How many passes did you set for the light cache? Are you hyper threading?

I set the LC Passes up to 16.

The Bios has a HT-Mode and this is enabled. The CPU runs with 2,53 Ghz @ 100%. It overclocked itself (CPU Feature).

In a production environment overclock does not make sense!

[3LP]

Hi there,
I know it's a bit off subjet, but I find this video on boxxtech site :

http://www.youtube.com/watch?v=3tDK9kj96Ck

I really wonder how this machine could get 51k point out off cinebench...
I have search the web to find some other persons that had this processors in there hands. I found that these people had achieve "only" 30k point with the platform Skulltrail and a dual QX9775 @ 4ghtz. The same as sown on the video.
http://forum.hardware.fr/hfr/Overclo...1.htm#t1614123

So what's the secret of boxxtech ? They have nearby the double of points with the same processors... ?

Does someone had a someday had a boxx like that in his hand and tried Vray Benchmark ? I think it could kick our ass, no ?

[Spikeadeliv]

x2 Xeons 5355s

Done on my work computer

x2 Xeon 5355s
8 Gb mem
Win XP 64
ATI FireGL V7600
Max 9
VRay SP4

Did a version on my OC'd i7 920 @ 3.2 Ghz (see 2nd Pix)
only 4 seconds slower then my work machine (see above).

[Lucetius]

So I've managed to get my hands on another super quick box. These chips aren't quite out yet and I've run in to a problem with them in that Max won't run with hyperthreading enabled but they are fast none the less.

I've run a couple different benchmarks through this and a couple other boxes including, my aging 8 processor, Opteron 280 box.

Results are below.

We'll start with the slowest box.
8 processor, AMD Opteron 280
Windows Advanced server 64
Vray 1.5 sp4




Intel E5530 2.4ghz Xeon. 4 cores per processor.
I ran test with and without HT on. While HT gave a slight boost, it was pretty negligable. I'm not showing the results here but I ran Cinebench and the results there were similar between HT on and off.

Here are the results with HT on.




X5560 2.6ghz Xeon 6 cores per processor
I could get Cinebench to run with/without HT but I could not get Max to run with HT on. These results are without HT. What's interesting is that the difference between results in Cinebench with/without HT was significant. There was a 20% increasse in speed with HT on. I'm bummed I can't test this with Max/Vray at the moment.
These processors have a lot of potential.

[3LP]

http://www.youtube.com/watch?v=BQ4sh...layer_embedded

outch

now what?

[AJ Jefferies]

We just got a couple of new i7 980x machines.

Here's the result from one of them.



Max 2011

[Da_elf]

anyone beaten 30 seconds yet?

[YoyoBoy]

My new machine arrives today or tomorrow. How can I get this scene? -or is a link for it that works buried somewhere in thread?

[YoyoBoy]

Nevermind, I found it.

[Fenikscg]

DELL Precision T7500
dual sixcore hyperthreaded X5680 @3,33Ghz
24gig RAM

Did I win something now ?

[sandykoufax]

Superfast indeed!

You win!

[vlado]

Hehe from 8 minutes to 30s, that's quite an improvement

Best regards,
Vlado

[percydaman]

yikes thats fast. Vlado what happened to that benchmark you were working on? I redownloaded this one the other day found on a few pages back. Man thats an old scene. Pre 'ALOT' of vray features. I even spent some time readjusting it for use with linear workflow/vray camera etc etc, in the aim of maybe rereleasing it, but I remembered your post awhile back about your own you were making.

[razzmus]

I think it would be super handy if one of the administrators was able to make a table with all these fine benchmark-results.
That way it would be waay easier to compare results and hardware.
Offcourse you would have to stick with only one scene.

I would suggest the following columns:
number of CPUs > CPU type > CPU clock > RAM amount > Operating System > Rendertime > System price in € > price/performance index