OctaneRender™ Standalone 3.03.2 Tweet

[zeigg]

Run any scene on 1080 and other older GPU and post results.

I don't know if you can consider this a "legit" benchmark.
I'm kind of new to benchmarking.
When I run the octane_22_benchmark.ocs scene on pathtracing I get between 5.35 - 5.41 Ms/sec.

This is using a single GTX 1080 Founder's Edition (From NVIDIA, not a partner card) with no overclocking or modifications.

[smicha]

Thank you.

[elihu252]

For what it's worth, I've been e-mailing with someone at Otoy and they said that at the moment the GTX 1080 puts about roughly the same performance as a single Titan BLACK. In other words it's not optimized at all atm. I just bought 2 1080s and I'm going to install them today to run my own non-scientific benchmarks. I'll post the results soon.

Working as a freelancer, I need to be very cost effective with all of the equipment I purchase. 1080s are theoretically the best tool for the job and I cannot wait for them to work properly in C4D. Otoy, light a fire under Nvidia's ass!

EDIT: Results from my very quick test can be found here viewtopic.php?f=11&t=55370

[zeigg]

For what it's worth, I've been e-mailing with someone at Otoy and they said that at the moment the GTX 1080 puts about roughly the same performance as a single Titan BLACK. In other words it's not optimized at all atm. I just bought 2 1080s and I'm going to install them today to run my own non-scientific benchmarks. I'll post the results soon.

Working as a freelancer, I need to be very cost effective with all of the equipment I purchase. 1080s are theoretically the best tool for the job and I cannot wait for them to work properly in C4D. Otoy, light a fire under Nvidia's ass!

EDIT: Results from my very quick test can be found here viewtopic.php?f=11&t=55370

I'm with you. Holding off to buy any more 1080's ATM, may consider waiting for the titan P or 1080 Ti's. Optimization likely have been done by then.

[coilbook]

Here is my post about a new bug viewtopic.php?f=9&t=55380
just in case it was not read there

[ramiz]

Ok each one took a while but... All at 16000 Max Samples:

DL = ~37 Ms/sec
PT = ~6 MS/sec
PMC = ~9.5 MS/sec

Here's a vid of me running it... unfortunately I didn't let it render all the way, I mostly wanted to test workflow feedback and not final render times.

[grimm]

Interesting results, definitely needs to be more optimized. Comparing your results to my 980 shows that your 1080 is much faster with the DL kernel and the PMC kernel. The 980 is still faster with the PT kernel.

DL: ~21 Ms/sec
PT: ~8.2 Ms/sec
PMC: ~5.4 Ms/sec

Did you turn off the alpha shadows? It's surprising to see that the PMC is faster, maybe because the 1080 is better with floating point calculations?

[smicha]

These are is exactly the same of my thoughts - maybe Marcus will tell us something more why PMC is that much faster? I looked at performance of my 780 6GB and a Titan and same results - DL and PMC are very fast for 1080...

[Elvissuperstar007]

very slowly I began to work open VDB ,
Alpha 3.0 is not so hampered
absorption inversion does not always work
ramp function creates a cube shell

[coilbook]

Please see crash errors non stop and 719 errors we never had in version 2 viewtopic.php?f=81&t=55402&p=283631#p283631