OTOY Forums

Hello, guys.

I run "start octane.exe --benchmark" (v3.00) and I would get better score in OctaneBench but...
Octane v3.00 isn't compatible with benchmark.orbx (v2.17) because the render is freeze at "Idea Info" scene.

Therefore, you should modify benchmark.orbx in v3.00. Or you should fix "--benchmark" argument.

Benchmark script fixed - OR 3.03.4 and no need scenes modified.

I got GTX 980 Ti SC+@3505MHz VRAM forced running Win 10 x64 build 14393 and [email protected] and I can end fine!

My result: Thank you!

nuno1980 wrote:Benchmark script fixed - OR 3.03.4 and no need scenes modified.

I got GTX 980 Ti SC+@3505MHz VRAM forced running Win 10 x64 build 14393 and [email protected] and I can end fine!

My result:

Octanebench 3.03.4_Win 10 [email protected]_GTX 980Ti SC+@mem 3505MHz forced-368.81WHQL (HQ).JPG

Thank you!

what do you mean your score is 147 almost
do you use only one 980ti ?
or you are rendering GPU + CPU ?
as far as I know that GPU + CPU is not implemented yet

MaaYaa wrote:what do you mean your score is 147 almost
do you use only one 980ti ?
or you are rendering GPU + CPU ?
as far as I know that GPU + CPU is not implemented yet

Yes, score is almost 147.
I use only a GTX 980Ti. We aren't worried despite 90%~95% GPU usage (for PMC kernel only with parallel samples 4).

nuno1980 wrote:

MaaYaa wrote:what do you mean your score is 147 almost
do you use only one 980ti ?
or you are rendering GPU + CPU ?
as far as I know that GPU + CPU is not implemented yet

Yes, score is almost 147.
I use only a GTX 980Ti. We aren't worried despite 90%~95% GPU usage (for PMC kernel only with parallel samples 4).

loool its even faster than 1080 .
well , if they dont optimize 1080 so it can hit 160 - 180 points , I dont see reason to upgrade towards pascal at all
it means no reason to buy pascal and pay 300$ more for 2GB of ram ,
+ 300$ for extra 10-15 points ? NO WAY
its like buying 6950x (paying almost 800$ for 2 additional cores )))))) in compare to 5960x , which costs 800-900$

2016 year hardware, software, development + prices = DISAPPOINTED !

MaaYaa wrote:...
2016 year hardware, software, development + prices = DISAPPOINTED !

I think it is a super year, but that's my point of view
“patience you must have my young padawan”

"We could narrow down the issue using the CUDA 8 RC toolkit and were able to work around it by doing a mixed CUDA 7.5/8 build for Pascal GPUs. For previous architectures we still use CUDA 7.5 only. This is all experimental and we haven't done any specific Pascal tweaks/optimizations yet and the performance isn't where we would expect it to be, but at least you can use your Pascal GPUs now."
quote from thread: viewtopic.php?f=24&t=55274

oh.. and the hbm2 will be out in near future... that might be a cheer up

I think a question as well is too what extent this new architecture hbm2 is optimizable for the octane engine and code, without sacrificing a lot on other architectures like kepler and fermi.

whersmy wrote:I think a question as well is too what extent this new architecture hbm2 is optimizable for the octane engine and code, without sacrificing a lot on other architectures like kepler and fermi.

If new GTX Titan with HBM2* or HBM3* will get normal performance then I'll buy it maybe in next year.

* - new VRAM type for 1 or 2 next-gen('s) of GTX Titan series only because new VRAM type will be very expensive.

Hey guys!

How to run or where did you get OctaneBench 3?

byMELT wrote:Hey guys!

How to run or where did you get OctaneBench 3?

If you're running Windows then copy "benchmark_data" (v2.17) folder and _"run_benchmark.bat" file to OctaneRender 3.x folder and run this batch file to benchmark ready. Good luck!