My Bench scene

[cecofuli]

Hi guy,

this is my scene for benchmark. the original scene is made for 3ds max 2008 and VRay 1.5 RC3.

. Antialiasing: Adaptive DMC 1 4
. GI : Brute Force + Brute Force 24 5
. Areashadows: 16 subdivs
. DOF : DMC 8 subdivs



SCENE DOWNLOAD

And here some result



That's the CODE

 
-----------------------------------------------------------------------
[b]AUTHOR[/b]:  [color="#2E8B57"][b]cecofuli[/b][/color]

[b]MACHINE[/b]: Laptop DELL M90

[b]Software[/b]: 3ds max 9 - 32bit

[b]Renderer[/b]:  VRay 1.5 RC3

[b]RAM[/b]: 2 GB

[b]CPU TEMP [/b]:  ???

[b]CPU THREAD [/b]:  2

[b]CPU[/b]: Core2Duo T7200  2000 @ 2000

[b]RENDETIME[/b]:  [color="#FF0000"] [b]1h 15m[/b][/color]
-----------------------------------------------------------------------

And that's a sample


-----------------------------------------------------------------------
AUTHOR: cecofuli

MACHINE: Laptop DELL M90

Software: 3ds max 9 - 32bit

Renderer: VRay 1.5 RC3

RAM: 2 GB

CPU TEMP : ???

CPU THREAD : 2

CPU: Core2Duo T7200 2000 @ 2000

RENDETIME: 1h 15m
-----------------------------------------------------------------------

If you like, this our thread, in Italian forum,

ORIGINAL THREAD

there are a lots of user with many configurations.


If you like, try to free use for your test and use a higher value for aa, brute force, etc... for fast machines.

Bye

cecofuli

[instinct]

Why bench with RC3 still when there's 1.5 and SP1 already?

Also the bench scene seems a bit limited as in WHAT it contains. Like no glossies, no caustics, no reflection, no refraction no fogged objects etc etc etc.

Kind Regards,
Thorsten

[cecofuli]

Oh, yes, becouse it's a old bench. I made this bench in march 2007 (You can see the date on the italian thread) .
And in the march 2007 the Dual Xeon QuadCore was very expensive.

This bench is only for workstation's bench purpose. For see only the rendertime (overclocking, speed test etc..) . Not for see the VRay's featueres.

bye

cecofuli

[instinct]

Sorry, didnt notice that it's so old. Still a bench scene should contain as many aspects as possible. For example some things arent effective on many cores as they are not good to be calced paralellised. So for real workd values a broad variety of features works better imho.

Kind Regards,
Thorsten

[cecofuli]

Oh, yes, much better with caustics, glossy, motion blur etc... but... sorry this is a my little free simply scene, only for testing CPU and rendertime in different CPU configuration (single machine and DR process)
In our site there are 32 page, so, perhaps you could interest to joint, or simply try it here.

Bye

cecofuli

[sv]

only for testing CPU and rendertime in different CPU configuration

I think what Thorsten is saying is that a benchmark scene should include many different rendering aspects (glossys, DOF, MB, SSS, reflections, refractions, AA etc)so as to truly give a rounded, thorough, accurate and balanced test.

[cecofuli]

Mmm... yes and not... I don't understand: what it means "accurate and balanced test"? if i have a lot of glossy refraction, SSS, MB etc... (in this scene i have DOF, bump, detailed textures, Areashadows, Adaptive AA, not so "light" scene...), the test is more correct?

For example, in this configuration, we see that a

(*) Dual Quad core x5355 2,66mhz = 14min 47,8sec LINK
(*) Q6600 @ 3600 OC = 24m 18sec LINK

This is the purpose= CPU rendertime. If I have more glossy reflection, ore Motion blur, i think that the Bench give me the same result. Much time, yes, but alwais correct.

Anyway, if you want try , the scene is here

Bye

cecofuli

[Janosik]

2 cecofuli :
maybe he means, when you have all this vray features on, that really shows how fast machines we have

BTW : can i ask for max7 version of this test ?