Iray Scene For Time Benchmarks by SickleYield

Daz Studio Public Beta 4.11
Stock scene as downloaded, unmodified. All times are to full completion.

GPU Only 1 x Geforce RTX 2070  = 2 minutes 21.14 seconds (Optix off)
GPU Only 1 x Geforce RTX 2070  = 1 minutes 49.11 seconds (Optix on)
GPU Only 1 x Geforce RTX 2070 + 1 x Geforce GTX 970 = 1 minutes 45.78 seconds (Optix off)
GPU Only 1 x Geforce RTX 2070 + 1 x Geforce GTX 970 = 1 minutes 19.47 seconds (Optix on)

All to full completion.  My 2070 is the MSI Armor, i.e. non-binned chip at stock.  When I benchmarked before with just the 970 I didn't get any benefit from Optix.  It made little to no difference.  With the 2070 it appears to.

Daz Studio Public Beta 4.11
Stock scene as downloaded, unmodified. All times are to full completion.

GPU Only 1 x Geforce RTX 2080 Ti = 1min 14.80sec (Optix on, spectral rendering off)
GPU Only 1 x Geforce RTX 2080 Ti = 1min 16.42sec (Optix on, spectral rendering on)
GPU Only 1 x Geforce RTX 2080 Ti = 1min 34.90sec (Optix off)

CPU Only Intel Core i7-8700K CPU(8th gen) 6 cores @ 3.70GHz CPU 16Gb ram (Optix On) = 20min 47.59sec

Both CPU & GPU = 1min 29.13sec (Optix on)

GPU only: 1 gtx1060 3gb + 1 gtx1060 6gb +2 gtx 970 4gd : 1 minute 14 seconds (Optix on)

GPU only: 1 gtx1060 3gb + 1 gtx1060 6gb +1 gtx 970 : 1 minute 35 seconds (Optix on)

GPU only: 1 gtx1060 3gb + 1 gtx1060 6gb + 1 gtx 560ti           : 2 minutes 14 seconds (optiX)
                1 gtx1060 3gb + 1 gtx1060 6gb                                : 2 minutes 29 seconds (optiX)

I7 - 6700K 4.00Ghz + 2 GTX 1070 MSI 1936Mhz

GTX1070 x 2 - 1m 48.s - 100% optix on - GPU only
GTX1070 x 2 - 1m 37.s - 100% optix on - GPU+CPU

GTX 1070 x 1 - 2m 52s - 100% optix on - GPU only
GTX1070 x 1 - 2m 42.s - 100% optix on - GPU+CPU

gpu oly on samsung notebook odyssey gp800 nbr1 i5 73000hq  geforce 1050 with latest drivers

2017-08-31 01:01:14.176 Finished Rendering
2017-08-31 01:01:14.212 Total Rendering Time: 8 minutes 39.2 seconds
2017-08-31 01:01:14.218 Loaded image r.png
2017-08-31 01:01:14.232 Saved image: C:\Users\xxxs\AppData\Roaming\DAZ 3D\Studio4\temp\RenderAlbumTmp\Render 1.jpg
2017-08-31 01:03:28.387 Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info : Device statistics:
2017-08-31 01:03:28.391 Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info : CUDA device 0 (GeForce GTX 1050):      4817 iterations, 0.498s init, 516.992s render

CPU: Dual quad core Xeon @ 2.26GHz
GPU: GTX 1070 Dual 
OptiX:  ON
3m 22.3s

CPU: Dual quad core Xeon @ 2.26GHz
GPU: GTX 1070 Dual 
OptiX: OFF
4m 36.3s

GPU: GTX1070 Dual
OptiX: On
2m 59.3s <----- Hu? The CPU slows it down?

GPU: GTX1070 Dual
OptiX: OFF
4m 47.4s

CPU: Dual quad core Xeon @ 2.26GHz
OptiX: ON
1h 20m 48.6s

CPU: Dual quad core Xeon @ 2.26GHz
OptiX: OFF
1h 5m 1.2s

GPU: GTX780M
OptiX: ON
11m 28.3s

CPU: i7 Quad @ 3.5 GHz
GPU: GTX780M
OptiX: ON
10m 19.2s <---- This CPU speeds things up, I guess the multithreading overcomes the payload of administrating the CPU.

CPU: i7 Quad @ 3.5 GHz
GPU: GTX780M
OptiX: OFF
23m 23.48

Conclusions
1. OptiX: +++ GPU rendering
2. OptiX: --  Old CPU rendering (Would be interesting to do the OptiX test on a newer CPU, maybe later)
3. Old CPUs: - GPU rendering
4. New CPU's: + GPU rendering
5. My GTX1070 Dual is awesome!

Let's hear it for old technology!

gtx 980 Ti x 2
gtx 780 Ti
Optix on

1 minutes 12.9 seconds

optix off

2 minutes 1.96 seconds


i7 4790/16 gig ram/GTX 980 Tix2/GTX 780 Ti/1 TB SSD/Coolermaster Haf X

Completed at 2 min 52 (CPU, GPU, GPU)
Completed at 2 mins 4 (GPU, GPU)
Completed at 36 minutes (CPU only) ----- owie!

System Specs:

Windows 10 Pro
i7-4930k @ 3.4 GHz
16gb RAM
Nvidia GTX 980Ti x2

When you say 90% do you mean setting the Rendering Converged Ratio to 90% or when the timer hits 90%? I didn't sit and stare at it, so I didn't see when it hit 90%. ;/ 

New Benchmark !

I7 - 6700k 4.00Ghz
Nvidia GTX 1070 MSI Gaming X 8Gb - OC mode - 1936 Ghz
32 Go RAM - 2133Mhz
Optix ON

GPU only - 90% in 23s - 100% in 2.52 min
CPU + GPU - 90% in 15s - 100% in 2.45 min

Hope it help !!

Huge difference than CPU only 90% in 21.16 min...

7 minutes, 34.41 seconds on an MSI Apache Pro laptop with nVidia GeForce GTX 1060, using DAZ Studio 4.0.3.166 release. I may try just CPU on this system (Intel Core i7-6700).

Daz Studio 4.9.3.117 Public Beta

AMD RX480...

NO MODIFIED SCENE - CPU only (I7 6700K - 4.0Ghz, 4 cores, 8 threads) : 21 minutes 16 seconds to reach 90%

I have been waiting for Pascal support for a couple of months and finally have it in a beta build so here goes.

Daz Studio 4.9.3.117 Public Beta

GPU: Asus Strix Gtx 1070 8Gb

90%:   2 minutes 20 seconds
100%: 3 minutes 22 seconds

Very happy with the results

Thanks SY for this benchmark scene. I downloaded it a while back but did not use it until now.

Machine:  Core i7 4790K 4.0 GHz not overclocked
               GPU NVIDIA GTX770 4GB
               32GB DDR3 951MHZ Ram

RESULTS:

GPU only with OptiX ON    6 minutes
GPU only with OptiX Off  10 minutes

CPU only  90% completed at 19 Minutes 30 Seconds
             100% completed at 24 minutes

CPU + GPU* with OptiX ON  5 minutes 40 seconds

*NOTE: CPU + GPU produces a lots of heat on both the CPU and GPU and increase fans speed and noise.
Using only CPU or GPU to render heats only the one in use. overall temperatures keep low.
I got a max temperature on the 80s centigrade.


I found the results very helpful. Thinking about upgrading my card.

Thanks

Thanks SY for providing us with this benchmark. Here's the result of my rig :

GPU Only (MSI GTX980ti) with OptiX Off : 3 minutes 50.27 seconds
GPU Only (MSI GTX980ti) with OptiX On : 2 minutes 38.79 seconds

Other spesifications :
Core i7 4790 @3.60GHz
VGEN 16GB DDR3

Still confused with OptiX Acceleration, since on some scenes it really speeds up the render, while on some others it really slows down the render.

Thanks! I've been debating about upgrading my graphics card, so this was a useful test! I've expanded the test a bit as well, as I've picked up tips for speeding up renders from other places, so decided this was a good place to test them. I'll put all my results up.

acer Aspire TC-150
15GB Ram
Windows 10
AMD Radeon HD7000

Render results (rendering to 80%)

Standard render (no modifications to scene) 15 mins

Modifications tests:
Bounding box changed to planes, one side left open - cancelled render at 69%, over 15mins

Bounding box changed to planes, one side left open, hdri map and extra light, fstop and shutter speed raised (apparently more light in a scene, and a place for light to escape leads to faster renders...we'll see....) holy moly, they weren't kidding! I was sceptical, but the results speak for themselves - 3 mins 25 seconds!

bounding box removed, back and floor left as background only, lights and camera as last test - not much change from previous test - 3 mins 31 

bounding box removed, back and floor left, lights and camera as in standard render - render stopped at 70% - over 15 mins 


Best result - increasing the amount of light and fstop to compensate. I don't think having the planes surrounding the woman and spheres actually made it much quicker, I think that was more to do with the heavy sss on spheres 8 and 9. I rarely render over 80% (I always render larger and then shrink) so this was a very useful test for me. Knowing that i can shave my render time from 15+ mins to just over 3 without having to splash out on a new graphics card simply by increasing the amount of light and adjusting the camera to compensate, just rendering on CPU, and rendering with memory optimised instead of speed - I don't think I'll splash out just yet.  Out of curiosity, I rendered the second to last test again with the spheres 8 + 9 removed: 1 MINUTE 50!!!

Intel Core i7-6700 CPU
16 GB RAM
Windows 10 - 64 Bit Operating System
GeForce GTX 970



Iray to 100% - CPU + GeForce = 3 Minutes 43 seconds
Iray to 100% - GeForce Only = 3 Minutes 57 Seconds

Going to be testing my laptop with this; the desktop is a complete waste of time to do the same.

First of all thank you. You are on the top of my favorites list for written and video tutorials. I have learned much. I mostly do 3Dlight renders with my laptop but have tried Iray. The renders timed out at 2 hrs, 4 hrs, and 6 hrs at 60 - 70% complete. At 12 hours it timed out at 98%. But that render was good enough to earn 3rd place in the freebie challenge. So with that in mind I decided to try your test render.

System  (Laptop)

Manufaturer    Toshiba
Model Satellite E55-A
Processor Intel Core i5-4200U CPU @ 1.6 GHz  2.30 GHz Dual Core 4 Thread
Installed RAM 6.00 GB
System Type 64 Bit
OS Windows Home 8.1

Iray Render Test 1 full scene
Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info : Maximum render time exceeded.
Total Rendering Time: 2 hours 5.5 seconds
rend info : CPU: 4941 iterations, 28.263s init, 7172.811s render

Iray Render Test 2 Orb 8 and 9 removed
Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info : Maximum number of samples reached.
Total Rendering Time: 1 hours 53 minutes 53.61 seconds
Loaded image r.png

Final results, I'm ordering a new computer in the morning.

Daz 4.9
I7 4770k ASUA z87 pro win 10 pro -64bit
32gig
2 evga gtx 770 4 gig (ebay used)
no overclock

90% 3 min 18 sec with GPUs... I don't know enough to do all the tests.

thank you this helps me understand my level of performance.

Benchmarked my new PC build

i7 6700K
16 GB RAM DDR4 2667mhz
GTX 980 Ti SC 6GB

Results:

GPU only: 2 minutes 18.73 seconds
CPU+GPU+OptiX Prime Acceleration: 2 minutes 22.23 seconds

Thank you for all of your work  SY

Giving it a go,

Combo of (1) GTX 970 and (1) 780ti Superclocked

CPU/GPU with OptiX Prime on (Intel Core i7-4770K 3.5 GhZ, 4 cores): 
2:30 to reach 90%

Actually have 2 GTX 970 but picked up a used 750ti because there are iray benchmarks showing it actually outperforming the Titan.  www.migenius.com/products/nvid…

I will post with all 3 cards when I find my PSU connectors.

Just thought I'd try this out.  (Optix accleration on)

results:

Finished Rendering
Total Rendering Time: 2 minutes 34.23 seconds


Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info : Device statistics:
Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info :     CUDA device 0 (GeForce GTX 980 Ti): 4593 iterations, 17.897s init, 134.912s render
Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info :     CPU (7 threads): 407 iterations, 12.478s init, 140.699s render

Not bad.

Thought I'd give give your scene another test not that I have some CUDAs to cook.

GTX 780ti = 3 mins 6 seconds
GTX 780ti + i7 4770 = 3 minutes 27 seconds
GTX 780ti (optix accelerator off) = 4 mins 40 seconds

Well, new system built both 980 sc cards running @ 16x on PCIE 3.0 render time for this benchmark 2 mins 4 sec @ 5000 iterations with Optix enabled. Pretty pleased with that. So it looks like running cards @ 16x gives much greater raw throughput

I just finished building my new system; Intel Core i7-5960X @ 3.00GHz, 8 cores. 32GB DDR4 memory; and Titan X GPU with 12GB RAM.

I haven't OC the system yet, but built it liquid cooled for the specific purpose of doing just that and tuning for rendering. As I am able to afford it, more Titan X cards will be put in the system.

GPU Only: 3 minutes 44 seconds

CPU + GPU: 3 minutes 14 seconds

CPU Only 14 min 50 seconds.

Wasn't able to figure out OC yet, but will do what I can. From reading the comments below I tried the OptX Prime Acceleration and the CPU+GPU render time dropped to less than a minute to 90%; 2 min 16 seconds to complete!

My MacBook Pro took over 20 minutes to complete.

From what I've been reading, adding more GPU (without SLI) will improve time in a pretty straightforward fashion - so I wonder what I could expect from two or more cards -

Great benachmark - for me GPU render only works best (2 x MSI GTX760, 2 GB, non-SLI mode) - render to 100% (=5000 iterations only) will take 4m20s with Optix enabled and 6m.30s without. 

Activating the CPU in my case slows down the system. 
Note: Update your Nvidia drivers (CUDA 3.0 might speed up things)

An interseting note here : I was running two 980gtx sc cards with a 3770k and 32gig of ram. As my cpu only supports 16 pcie lanes the two cards ran @ 8x/8x. What I found was the when i removed one card so the remaining card was running at full 16x rendering times were smashed. This scene for example I can complete in between 3 and 4 minutes. Am upgrading to a 5930k soon which will allow both gpus to run @ x16 as there are 40 pci lanes on this cpu. Your cards pcie lane speed  seems to be crucial to using high end cards to their full potential. So in summary having 2 or more cards will become less productive if they are bottlenecked by their throughput. Hope this helps.
 

Benchmarked my new build

i7 4930K
32GB RAM
GTX 970 4GB

CPU+GPU+OptiX Prime Acceleration: 3mins 35secs.
GPU+OptiX Prime Acceleration: 3mins 41secs
CPU+GPU: 5mins 20secs
GPU Only: 5mins 48secs
CPU Only: 25mins 31secs

It would seem that the CPU barely makes a dent in the times, despite being a 6 core i7, and the Optix Prime Acceleration makes quite a difference. 

Im running this benchmark in Windows 7.
I wanted to run it in OSX 10.10. but I keep getting the error: 

WARNING: /src/pluginsource/DzIrayRender/dzneuraymgr.cpp(256): Iray ERROR - module:category(IRAY:RENDER):  1.3   IRAY   rend error: CUDA device 0 (GeForce GTX 970): an illegal instruction was encountered (while launching CUDA renderer)
WARNING: /src/pluginsource/DzIrayRender/dzneuraymgr.cpp(256): Iray ERROR - module:category(IRAY:RENDER):  1.3   IRAY   rend error: CUDA device 0 (GeForce GTX 970): Failed to launch renderer
WARNING: /src/pluginsource/DzIrayRender/dzneuraymgr.cpp(256): Iray ERROR - module:category(IRAY:RENDER):  1.2   IRAY   rend error: CUDA device 0 (GeForce GTX 970): an illegal instruction was encountered (while de-allocating memory)

Not sure why its doing that, it only seems to happen with the default bree texture and with a few other textures and shaders. Hope DAZ get it fixed !

Thank you

Complete scene to 100%
GPU only (Nvidia GTX980 4GB x2): 2 minutes 4 seconds