[Maxer]

I ran my test last night using 8 threads on my Dual Opteron 248 and my time went from 1hr 19min 27sec to 44min 23sec

That's an 80% increase in speed

[andronikos916]

happy that you gained speed MAxwer...

cy,
A

[andronikos916]

sure... go ahead and make a nice new graph. to busy testing other stuff at the moment, sorry guys.

cy,
A

[Steve Casey]

Hi Andronikos,

Just rendered the skydome.zip benchmark scene using if these are any good for you.

Dual xeon 3.2Ghz, 2gig ddr266 ram,
maxwell version 1_1_30_0001_alpha,
4 threads
Time 38m 28s.

Then tried it with maxwell version 1_1_31_0001_alpha, to see if there was any imrovement in the render time, same configuration as above with a render time of: 38m 33s.

see ya
steve

[thomas lacroix]

i find odd that the dual xeon 3.2 seems slower than the 2.8 so as the opteron 248...

are the graphics updated?
news anyone ? did someone try an dual opteron 250 or a dual xeon 3.6 ?

[Maxer]

I'm guessing the dual Opteron 248 is newer than the Xeon 3.2, that's just a guess. If not then I'm glad I've got the Opteron cause it's kicking the hell out of Intel.

[j_petrucci]

Here is my test with Pentium IV 3.2Ghz, Extreme Edition 2MB cache

Code: Select all

[1h 54m 7s] of [16h 39m 0s] , current sampling level : 15.04 

Render finished successfully :) 

so, my elapsed time is 1h54m7s with Maxwell Render 1.2.2a

PS: andronikos, your MXS is set up to work with Maxwell in "Program Files" folder (I have it under "Programmi"):

[pluMmet]

Here is my test with Pentium IV 3.2Ghz, Extreme Edition 2MB cache

Code: Select all

[1h 54m 7s] of [16h 39m 0s] , current sampling level : 15.04 

Render finished successfully :) 

So in the newest version of M~R a Pentium 3.2 takes twice as long as a pentium 2.8 in the old 1_1_10_0001_alpha_

[tom]

No pluMet, sampling level is not the same as in the past.

[pluMmet]

No pluMet, sampling level is not the same as in the past.

ok sorry...
well I've now done the scene in 1.2.2a

I did it on one of the nodes. I'm supprised that my dual amd 2000+ is (so much) faster then a 3.2 EE

[slobodan]

Hi guys,

Not to be mean or anything but I think that Ignacio said use Maxwellbench in the log file for benchmarking.

Just my two cents.

Slobodan

[andronikos916]

You can use also this one... no problem I beleive. Just include at your image the MAxwell version and PC sprecs plz

cy,
Andronikos

[LarsSon]

Hi

IS THIS NEW RECORD? MaxwellBench 1.0: 85.074692

I just got my new workstation and tested it with skydome.mxs.
I this is really good result. What do you think.



MAXWELL 1_2_2a_beta ( Engine core: rs1c )
[ Physically and Unbiased Based Rendering ]


## Scene Information ##

44 meshes ( 37798 triangles )

## Render Information ##

*settings:

resolution: 640 x 480

*render time: 0 hours 33 minutes 19 seconds

target sampling level: 15
sampling level reached: 15

MaxwellBench 1.0: 85.074692

-LarsSon

[slobodan]

Larson,

Your result is mad...
What did you spend, $10000 for that maschine??

Slobodan

[LarsSon]

I know it's mad! I just got this new machine and i just can't
stop smiling these days.

Actually 15000€ Well it was companys money, but
i'm very happy now. And i'm going to explode next month,
when i'll have five more to be render slaves.


-LarsSon