I have made some test with the new v2.0 and all my renders are slower than the v1.55 from 15% to 25%.
This is a serious drop down performance, what's changed ?
Why v2.0 so slow ?
Forum rules
Please post only in English in this subforum. For alternate language discussion please go here http://render.otoy.com/forum/viewforum.php?f=18
Please post only in English in this subforum. For alternate language discussion please go here http://render.otoy.com/forum/viewforum.php?f=18
This is the price of all new feature. Unfortunately every new feature slows down gpu code even if it is not used. Not sure we can improve the speed anytime soon.Dom74 wrote:I have made some test with the new v2.0 and all my renders are slower than the v1.55 from 15% to 25%.
This is a serious drop down performance, what's changed ?
It would be nice if we could have 1.55 and 2.0 switchable I guess then - as from my tests we have to copy DLL's around for now or have multiple 3dsmax installations.
It's got to be said that given all the 2.0 goodies present and planned then really the thing that becomes more important now is speed to me personally - not that it's a slouch at all and I understand the cloud based direction and reliance on future GPU scaling etc.
Thanks for the great work on 2.0 and I am sorry we always ask for more
It's got to be said that given all the 2.0 goodies present and planned then really the thing that becomes more important now is speed to me personally - not that it's a slouch at all and I understand the cloud based direction and reliance on future GPU scaling etc.
Thanks for the great work on 2.0 and I am sorry we always ask for more

_________________
RTX 3090, Ryzen 5950x, 32GB RAM, Windows 10 64-bit
RTX 3090, Ryzen 5950x, 32GB RAM, Windows 10 64-bit
WOW, almost 50% slower renders here... !
AMD 9950x3D / RTX 4090 / RTX 3090 / 192 Go DDR5 / Win 11
http://www.remy-kerbiquet.com
http://www.remy-kerbiquet.com
I guess it depends what features You use =) curious though if the noise meassured at teh same sample level looks the same =) as I can remmeber a reease where it was not so fast (but noice looked better at the same amount of samples)remKa wrote:WOW, almost 50% slower renders here... !
I just guess optimisation will come as features will be implimented and code will be polished out =) noce to see v2 out from the door =)
Hi,Karba wrote:This is the price of all new feature. Unfortunately every new feature slows down gpu code even if it is not used. Not sure we can improve the speed anytime soon.Dom74 wrote:I have made some test with the new v2.0 and all my renders are slower than the v1.55 from 15% to 25%.
This is a serious drop down performance, what's changed ?
I hope you can, I understand news feature can slow down a bit, but 25% loss (and even more) is a big loss.
Win7 x64 - 3DSMAX 2019 - 32GB RAM - 1080 TI


----------------------------------------------------------------------------------------------------------------------------------------
intel core i7-4960x 3.60 ghz
32 gb ram
win 7 64 bit
1x nvidia geforce gtx 780
1x nvidia geforce GTX TITAN
intel core i7-4960x 3.60 ghz
32 gb ram
win 7 64 bit
1x nvidia geforce gtx 780
1x nvidia geforce GTX TITAN
- prehabitat
- Posts: 495
- Joined: Fri Aug 16, 2013 10:30 am
- Location: Victoria, Australia
Anyone with a Fermi & Kepler report the difference between ~1.50 & 2.0?
I'm interested to know if the older CUDA implementation & hardware architecture tests differently
I'm interested to know if the older CUDA implementation & hardware architecture tests differently
Win10/3770/16gb/K600(display)/GTX780(Octane)/GTX590/372.70
Octane 3.x: GH Lands VARQ Rhino5 -Rhino.io- C4D R16 / Revit17
Octane 3.x: GH Lands VARQ Rhino5 -Rhino.io- C4D R16 / Revit17
Is SLI disabled? We have seen a slow down in the range 10-20% compared to 1.5, maybe 25%, but not 50%. Could you post / send us the scene and the actual speeds you measure? Thanks.criko3d wrote:this is way very slow.... i honestly expected more of this... in here is nearly 50% slower on dual titans..
In theory there is no difference between theory and practice. In practice there is. - Yogi Berra
- gabrielefx
- Posts: 1701
- Joined: Wed Sep 28, 2011 2:00 pm
do you want my 400Mb scene with 118 active textures to do a comparison?
The performance loss is exponential.
The performance loss is exponential.
quad Titan Kepler 6GB + quad Titan X Pascal 12GB + quad GTX1080 8GB + dual GTX1080Ti 11GB