Daz Studio Iray - Rendering Hardware Benchmarking

1246

Comments

  • contactcontact Posts: 11

    there is a problem  : it seems we are not using rt coresbecause if you compare the speed of the dual 2080ti withIray RTX 2019.1.3 you get 14.5 it/s and with 1.3 Iray 2018.1.3 you get 11.8 it/s that only a 22% speedup so we are far from the 10x speedup rt cores are supposed to give above cuda cores. has someone an explanation for that ? thanks

  • contactcontact Posts: 11

    System Configuration
    System/Motherboard: hp omen 15 laptop
    CPU: core i5-9300h
    GPU: rtx2060
    System Memory: 2x8gb
    OS Drive: samsung 850 256gb
    Asset Drive: same
    Operating System: win10 home 1809
    Nvidia Drivers Version: studio driver 431.86
    Daz Studio Version: 4.12.086
    Optix Prime Acceleration: on

    Benchmark Results : 3.5 iterations per second CUDA device 0 (GeForce RTX 2060):      1800 iterations, 0.459s init, 513.598s render

  • contactcontact Posts: 11

    fyi "optix prime acceleration" checked or not checked makes no difference on my rtx2060 laptop ( in the log i have "IRAY   rend warn : The 'iray_optix_prime' scene option is no longer supported.")

  • RayDAntRayDAnt Posts: 584
    edited October 2019
    contact said:

    there is a problem  : it seems we are not using rt coresbecause if you compare the speed of the dual 2080ti withIray RTX 2019.1.3 you get 14.5 it/s and with 1.3 Iray 2018.1.3 you get 11.8 it/s that only a 22% speedup so we are far from the 10x speedup rt cores are supposed to give above cuda cores. has someone an explanation for that ? thanks

    Here's how it works: RTCores ONLY speed up raytracing - which is but a single part of the overall 3D rendering process. Each indivudal scene will have its own ratio of raytracing to non-raytracing workload to it based primarily on geometry complexity, density and lighting effects used in that scene. And since the benchmarking scene around which this thread is based happens to be on the lighter side where raytracing workload is concerned (small memory footprint scenes are kind of that by definition) the speedup observed is on the lower end of the expected range. If you want to see examples where the speedup from RTCore utilization is much more apparent, see this other benchmarking thread and particularly this post where the largest speedup observed (by me, actually) currently stands at 11.87x or 1,187%.

    Post edited by RayDAnt on
  • outrider42outrider42 Posts: 2,194
    Yes, it is highly dependent on the scene and can vary dramatically. But the general take away is this: the more geometrically complex the scene is, the more of an impact the RT cores will have, making renders far faster than without.

    Also OptiX has completely changed in the new Iray, which is why the setting makes no difference. That toggle is basically a legacy item, and will likely be removed at some point in future releases. Note the toggle refers to "OptiX Prime", the new Iray uses full OptiX, they are not the same thing.
  • RayDAntRayDAnt Posts: 584
    edited October 2019

    Fyi there's a new version of Iray out.

     

    System Configuration
    System/Motherboard: Gigabyte Z370 Aorus Gaming 7
    CPU: Intel i7-8700K @ stock (MCE enabled, watercooled)
    GPU: Nvidia Titan RTX @ stock (watercooled)
    System Memory: Corsair Vengeance LPX DDR4 32GB @ 32000Mhz
    OS Drive: Samsung Pro 970 NVME SSD 512GB
    Asset Drive: Sandisk Extreme Portable SSD 1TB
    Operating System: Windows 10 1903
    Nvidia Drivers Version: 436.48
    Daz Studio Version:  4.12.1.016 Beta x64

    Benchmark Results - Titan RTX (TCC mode)
    Total Rendering Time: 3 minutes 48.31 seconds
    CUDA device 0 (TITAN RTX): 1800 iterations, 2.117s init, 223.662s render
    Iteration Rate: (1800 / 223.662) = 8.048 iterations per second
    Loading Time: ((180 + 48.31) - 223.662) = 228.31 - 223.662 = 4.648 seconds

    Benchmark Results - Titan RTX (WDDM mode)
    Total Rendering Time: 3 minutes 54.66 seconds
    CUDA device 0 (TITAN RTX): 1800 iterations, 2.278s init, 229.775s render
    Iteration Rate: (1800 / 229.775) = 7.834 iterations per second
    Loading Time: ((180 + 54.66) - 229.775) = 234.66 - 229.775 = 4.885 seconds

    Benchmark Results - i7-8700K
    Total Rendering Time: 1 hours 2 minutes 3.25 seconds
    CUDA device 0 (TITAN RTX): 1800 iterations, 2.386s init, 3718.209s render
    Iteration Rate: (1800 / 3718.209) = 0.484 iterations per second
    Loading Time: ((3600 + 120 + 3.25) - 3718.209) = 3723.25 - 3718.209 = 5.041 seconds

     

    System Configuration
    System/Motherboard: Microsoft Surface Book 2
    CPU: Intel i7-8650U @ stock
    GPU: Nvidia GTX 1050 2GB @ stock
    System Memory: 16GB DDR3 @ 1867Mhz
    OS Drive: Samsung OEM 512GB NVME SSD
    Asset Drive: Sandisk Extreme 1TB External SSD
    Operating System: W10 version 1903
    Nvidia Drivers Version: 436.48 WDDM

    Benchmark Results - GTX 1050 2GB
    Total Rendering Time: 38 minutes 55.61 seconds
    CUDA device 0 (GeForce GTX 1050): 1800 iterations, 5.040s init, 2327.177s render
    Iteration Rate: (1800 / 2327.177) = 0.773 iterations per second
    Loading Time: ((2280 + 55.61) - 2327.177) = 2335.61 - 2327.177 = 8.433 seconds

    Post edited by RayDAnt on
  • pctech4nypctech4ny Posts: 102
    edited October 2019

    EDIT: added 2nd run with Optix enabled.

    If I did it right I ran a benchmark test of my new GTX 1660 since it seems nobody has reported one yet.

    System Configuration
    System/Motherboard: MSI B75A-G43
    CPU: Intel Core i7-3770 3.40GHz
    GPU: Gigabyte GTX 1660 6GB
    System Memory: Crucial 32 GB DDR3
    OS Drive: WD Blue 1TB
    Asset Drive: WD Black 2TB
    Operating System: Windows Pro 7 64-bit build 7601
    Nvidia Drivers Version: 436.15 standard
    Daz Studio Version: 4.11.0.383 x64
    Optix Prime Acceleration: OFF

    Benchmark Results
    DAZ_STATS

    2019-10-18 21:15:32.506 Finished Rendering
    2019-10-18 21:15:32.538 Total Rendering Time: 13 minutes 29.71 seconds


    IRAY_STATS

    2019-10-18 21:16:53.332 Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info : Device statistics:
    2019-10-18 21:16:53.333 Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info : CUDA device 0 (GeForce GTX 1660):      1800 iterations, 6.516s init, 799.269s render


    Iteration Rate: (DEVICE_ITERATION_COUNT / DEVICE_RENDER_TIME) 2.252 iterations per second


    Loading Time: ((TRT_HOURS * 3600 + TRT_MINUTES * 60 + TRT_SECONDS) - DEVICE_RENDER_TIME) 10.441 seconds

    ***

    2nd run with Optix enabled:

    Optix Prime Acceleration: ON

    Benchmark Results
    DAZ_STATS

    2019-10-19 09:34:38.101 Finished Rendering
    2019-10-19 09:34:38.122 Total Rendering Time: 12 minutes 6.70 seconds

    IRAY_STATS


    2019-10-19 09:35:22.618 Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info : Device statistics:
    2019-10-19 09:35:22.618 Iray INFO - module:category(IRAY:RENDER):   1.0   IRAY   rend info : CUDA device 0 (GeForce GTX 1660):      1800 iterations, 5.229s init, 717.052s render

    Iteration Rate: (DEVICE_ITERATION_COUNT / DEVICE_RENDER_TIME) 2.510 iterations per second
    Loading Time: ((TRT_HOURS * 3600 + TRT_MINUTES * 60 + TRT_SECONDS) - DEVICE_RENDER_TIME) 9.648 seconds

    benchmark run 1 GTX 1660 no OptiX.png
    900 x 900 - 1M
    benchmark run 2 GTX 1660 with OptiX.png
    900 x 900 - 1M
    Post edited by pctech4ny on
  • RayDAntRayDAnt Posts: 584

    @pctech4ny Yep, that's it!

  • pctech4nypctech4ny Posts: 102
    RayDAnt said:

    @pctech4ny Yep, that's it!

    I just added a 2nd run with OptiX enabled and added to my previous post.

    (I think I'd better  leave that enabled from now on...)

  • System Configuration
    System/Motherboard: Asrock B450M Steel Legend
    CPU: AMD Ryzen 5 2600 @ stock
    GPU: Asus GTX 1050 ti Cerberus @ stock
    System Memory: Corsair Vengence 16Gb @ 3200mhz
    OS Drive: Intel 660p m.2 nvme 512GB
    Asset Drive: SanDisk Ultra II 240GB
    Operating System: Windows 10 pro 1903 18362.418
    Nvidia Drivers Version: 431.60
    Daz Studio Version: 4.12.0.86

    Benchmark Results
    Optix Prime Acceleration: off
    Total Rendering Time: 29 minutes 29.59 seconds

    Iteration Rate: 1.017 iterations per second
    Preload Time: 8.134 seconds

     

    Bench GPU only stock.jpg
    900 x 900 - 475K

  • System/Motherboard: Asrock B450M Steel Legend
    CPU: AMD Ryzen 5 2600 @ stock
    GPU: MSI RTX 2060 Super Ventus @ stock
    System Memory: Corsair Vengence 16Gb @ 3200mhz
    OS Drive: Intel 660p m.2 nvme 512GB
    Asset Drive: SanDisk Ultra II 240GB
    Operating System: Windows 10 pro 1903 18362.418
    Nvidia Drivers Version: 441.12
    Daz Studio Version: 4.12.0.86

    Benchmark Results
    Optix Prime Acceleration: off
    Total Rendering Time: 6 minutes 50.6 seconds

    Iteration Rate: 4.383 iterations per second
    Preload Time: 7.065 seconds
     

    rtx 2060 super.png
    900 x 900 - 1M
  • outrider42outrider42 Posts: 2,194

    Thanks for posting @droidy001. Finally a 2060 Super test. The 2060 Super can beat a 1080 and even knock on 1080ti and Titan XP territory. That proves just how good of a deal the 2060 Super is.

    By comparison, the 1660 lags quite a bit behind. That is quite a gap between them.

  • RayDAntRayDAnt Posts: 584

    @droidy001 was the device render time in the log file for your 2060 SUPER 403.535s? Just wanted to double-check since your results appear so much better than the 2070 (granted the only numbers we have for the 2070 right now are from Iray pre-RTX.)

  • outrider42outrider42 Posts: 2,194

    You can double check by calculating the reported iterations per second at 1800 iterations. I get 6.8446 minutes, which comes out to 6 minutes and 50 seconds, matching the time. So it looks to be the correct time to me. And with RT cores it makes sense to me, because that is the only way a 2060 could be so close to a 1080ti or Titan XP.

  • RayDAnt said:

    @droidy001 was the device render time in the log file for your 2060 SUPER 403.535s? Just wanted to double-check since your results appear so much better than the 2070 (granted the only numbers we have for the 2070 right now are from Iray pre-RTX.)

    I didn't save the log, so I will run it again and attach it. You can make sure I'm reading it correctly
  • RayDAntRayDAnt Posts: 584
    edited November 2019

    You can double check by calculating the reported iterations per second at 1800 iterations. I get 6.8446 minutes, which comes out to 6 minutes and 50 seconds, matching the time. So it looks to be the correct time to me. And with RT cores it makes sense to me, because that is the only way a 2060 could be so close to a 1080ti or Titan XP.

    Given that @droidy001 reported a Total Rendering Time of 6 minutes 50.6 seconds (410.6 seconds), an Iterations Per Second of 4.383, and a Loading Time of 7.065 seconds along with the fact that Loading Time is calculated as:

    Loading Time = Total Rendering Time - Device Rendering Time

    It follows that:

    7.065 = 410.6 - Device Rendering Time
    Device Rendering Time + 7.065 = 410.6
    Device Rendering Time = 403.535

    And given that the benchmark scene runs to 1800 Total Iterations and Iteration Rate is calculated as:

    Iteration Rate = Total Iterations / Device Rendering Time

    If also follows that:

    Iteration Rate = 1800 / 403.535
    Iteration Rate = 4.461 (not 4.383!)

    Meaning that there is a discrepancy somewhere in these numbers.

    Post edited by RayDAnt on
  • Here is the log

    txt
    txt
    log.txt
    122K
  • RayDAntRayDAnt Posts: 584
    droidy001 said:

    Here is the log

    Total Rendering Time: 6 minutes 50.98 seconds
    [...]
    1800 iterations, 2.562s init, 405.053s render

    Got it. Thanks!

    So that would give:

    Iteration Rate: (1800 / 405.053) = 4.444 iterations per second
    Loading Time: ((6 * 60 + 50.98) - 405.053) = (410.98 - 405.053) = 5.927 seconds

  • outrider42outrider42 Posts: 2,194

    Well, that makes the iteration per second even higher, LOL.

  • RayDAntRayDAnt Posts: 584
    edited November 2019

    Well, that makes the iteration per second even higher, LOL.

    Yeah, it does. Mayhap we can convince someone with a 2070 to run the benchmark in 4.12 so we can see if ithe 2060 SUPER truly beats out one of its higher tier brethren (if so, that would be a first across the entire Turing range - at least in terms of Iray performance.)

    Post edited by RayDAnt on
  • droidy001droidy001 Posts: 69
    edited November 2019

    Well, that makes the iteration per second even higher, LOL.

    Definitely seems like value for money then when compared with other cards. If it's of any interest I paid £380. The same store have their last remaining model of the 1080 up for £540
    Post edited by droidy001 on
  • outrider42outrider42 Posts: 2,194

    Indeed. I know a lot of people don't like RTX prices, and in many ways they would be right. But the 2060 and especially the 2060 Super are actually quite good for what they offer. You are not going to beat that without going into the used market, and even then it would be tough as the 1080 still sells high used, and it isn't even as fast...and now that the 2060 Super matches the 1080 with 8GB, its almost no contest. 

    I figured I give 1.3 a go. This time I tested the 1080tis alone to see how they compare to a 2060 Super, and then ran them together.

    CPU: i5 4690K

    GPU #1:  EVGA 1080ti SC2

    GPU #2: MSI 1080ti Gaming

    RAM 32GB HyperX

    OS Drive Samsung 860 EVO 1TB

    Asset Drive: Samsung 860 EVO 1TB and WB 4TB Black HDD

    Driver 436.02

    Daz 4.12.0.86

     

    MSI 1080ti ONLY

    2019-11-09 00:05:22.782 Total Rendering Time: 7 minutes 42.23 seconds

    2019-11-09 00:05:50.885 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : Device statistics:

    2019-11-09 00:05:50.885 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 1 (GeForce GTX 1080 Ti): 1800 iterations, 6.270s init, 452.016s render

    EVGA 1080ti ONLY

    2019-11-09 00:22:21.167 Total Rendering Time: 7 minutes 38.53 seconds

    2019-11-09 00:22:28.149 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : Device statistics:

    2019-11-09 00:22:28.149 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 0 (GeForce GTX 1080 Ti): 1800 iterations, 5.600s init, 449.931s render

     

    Both cards

    2019-11-09 00:13:08.008 Total Rendering Time: 4 minutes 2.43 seconds

    2019-11-09 00:13:17.099 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : Device statistics:

    2019-11-09 00:13:17.099 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 0 (GeForce GTX 1080 Ti): 911 iterations, 6.063s init, 233.354s render

    2019-11-09 00:13:17.099 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 1 (GeForce GTX 1080 Ti): 889 iterations, 6.299s init, 233.061s render

    After this test I decided to update my Nvidia driver

    Driver 441.12

    Both

    2019-11-09 00:53:12.046 Total Rendering Time: 4 minutes 0.1 seconds

    2019-11-09 00:53:27.143 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : Device statistics:

    2019-11-09 00:53:27.143 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 0 (GeForce GTX 1080 Ti): 904 iterations, 6.463s init, 229.385s render

    2019-11-09 00:53:27.149 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 1 (GeForce GTX 1080 Ti): 896 iterations, 6.924s init, 228.953s render

     

    Both +50 Overclock

    2019-11-09 01:00:22.623 Total Rendering Time: 3 minutes 56.39 seconds

    2019-11-09 01:00:29.163 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : Device statistics:

    2019-11-09 01:00:29.163 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 0 (GeForce GTX 1080 Ti): 901 iterations, 5.858s init, 226.760s render

    2019-11-09 01:00:29.163 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 1 (GeForce GTX 1080 Ti): 899 iterations, 6.278s init, 227.107s render

     

    So...a 1080ti LOSES to a 2060 Super. That is simply incredible by any standard. The top gaming card (if you don't count the Titans) from 2017 has lost to a card that is only a 4th tier level in 2019. I do not think this has ever happened in a single generation. Granted it is at a specific task, but that does not change what an amazing thing this is. I cannot recall any instance of a x60 beating the x80ti from the previous generation at anything, nor the x80 from before they started adding "ti" to models. Holy cow.

    But I did tell you guys this was going to be possible with RTX! So while I am kind of saddened that my sexy EVGA and MSI 1080ti can lose to some scrawny 2060 Super that can't even fill out its shroud (seriously, I bet that 2060 uses padding...), I am also quite excited for the future.

    At least I can still comfort myself by their combined power, and with their powers combined they form a Mighty Megazord of rendering goodness and pack 11GB of Zordon's VRAM. But it appears that is the only way they can win. Perhaps I should upgrade my 1080tis to 2060 Supers, LOL.

  • DripDrip Posts: 410
    edited November 2019

    System Configuration
    System/Motherboard: MSi B450-A Pro
    CPU: AMD Ryzen 5 2600x stock
    GPU: RTX2070 @ stock
    System Memory: Corsair LPX 8GB @ default
    OS Drive: Samsung EVO 860 500GB
    Asset Drive: Seagate Barracuda 4TB @ 5400rpm
    Operating System: Win 10 Home
    Nvidia Drivers Version: 441.12 Studio
    Daz Studio Version: 4.12.0.86
    Optix Prime Acceleration: On

    Benchmark Results
    DAZ_STATS    6 minutes 49.16 seconds
    IRAY_STATS
    Iteration Rate: 4.5600559
    Loading Time: 14.428


    2019-11-09 08:55:43.038 Finished Rendering
    2019-11-09 08:55:44.491 Total Rendering Time: 6 minutes 49.16 seconds
    2019-11-09 08:55:49.054 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : Device statistics:
    2019-11-09 08:55:49.061 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 0 (GeForce RTX 2070):      1800 iterations, 7.579s init, 394.732s render

     

    (including the lines from the log, just in case I messed up calculations again)

    Post edited by Drip on
  • outrider42outrider42 Posts: 2,194
    Its not surprising at all. All the gaming benchmarks found the 2060 Super neck and neck with the original 2070, if only marginally slower. Like less than 5%. And that was a stock model, overclocking shrank that gap more. A 3rd party card may be clocked higher. The card seen here boosts to 1665, which is a very minor boost over vanilla, but hey, its still a boost. But this comes down to more than simple boost clocks. The MSI card almost certainly has a better cooler as well, and due to temp performance, it can maintain a higher boost for longer.

    Nvidia turned its whole RTX product stack on its ear with the Super lineup. Its more of a mid generation refresh almost a year after the originals, so not so far fetched. Its just a really goofy naming scheme.

    The 2070 still bets it. Not by much at all, but a 10 second margin is a clear win.
  • @outrider as you mention this is task specific. Things may be a lot different in a gaming environment. Not being a gamer I couldn't comment. What I can do though is run a Timespy bench.
  • RayDAntRayDAnt Posts: 584

    Thanks @Drip!

    @outrider42 are you still on Windows 10 1809?

  • outrider42outrider42 Posts: 2,194
    edited November 2019
    I already know that answer. The 2060 Super wont beat a 1080ti in games...unless that game uses real time ray tracing. Then it absolutely will. However the 2060 and 2060 Super have the fewest number of RT cores in the RTX line, so their ray tracing performance in games isn't that great. It depends on the game, but in most cases any ray tracing features and/or resolution must be toned down in order to be playable (generally considered to be a minimum 60 fps). In some games you can hit that, others you cannot.

    What that translates to is that because the ray tracing of the 2060 cards is a bit limited, many gamers would ignore using the ray tracing. Thus that goes back to standard rasterization. Here the 2060 and Super are both still solid. They can match or get close to a 1080 in most games, which is still a big leap. But they cannot touch the 1080ti. Even the 2070 Super can't beat the 1080ti in rasterization most of the time, though it can win in some games.
    Post edited by outrider42 on
  • outrider42outrider42 Posts: 2,194
    RayDAnt said:

    Thanks @Drip!

    @outrider42 are you still on Windows 10 1809?

    No, its 1903.
  • amadeus00amadeus00 Posts: 8
    edited November 2019

    System Configuration
    System/Motherboard: Asus X99-A II
    CPU: Intel i7-6850K @ 3.8GHz
    GPU: GPU1 Asus 980TI OC @ 1317MHz GPU eVGA 2080TI @ 1650MHz
    System Memory: G.Skill DDR4 [email protected]
    OS Drive: Samsung PM981 Polaris 1TB
    Asset Drive: BRAND MODEL CAPACITY/Same (if same as OS)
    Operating System: Windows 10 Pro 1903
    Nvidia Drivers Version: 441.12
    Daz Studio Version: 4.12.0.86
    Optix Prime Acceleration: n/a

    Benchmark Results - GTX 980Ti OC
    Total Rendering Time: 10 minutes 42.35 seconds
    CUDA device 0 (GeForce GTX 980 Ti):      1800 iterations, 3.071s init, 634.380s render
    Rendering Performance: 1800 / 634.380s = 2.837 iterations per second
    Loading Time: (600s + 42.35s) - 634.380s = 7.97 seconds

    Benchmark Results - GTX 980Ti OC + GTX 980TI
    Total Rendering Time: 5 minutes 44.9 seconds
    CUDA device 1 (GeForce GTX 980 Ti):      861 iterations, 4.067s init, 332.835s render
    CUDA device 0 (GeForce GTX 980 Ti):      939 iterations, 3.726s init, 333.303s render
    Rendering Performance: 1800 / 333.303s = 5.219 iterations per second
    Loading Time: (300s + 44.9s) - 333.303s = 11.597 seconds

    Benchmark Results - RTX 2080Ti
    Total Rendering Time: 4 minutes 20.86 seconds
    CUDA device 0 (GeForce RTX 2080 Ti):      1800 iterations, 8.463s init, 248.357s render
    Rendering Performance: 1800 / 248.357s = 7.248 iterations per second
    Loading Time: (240s + 20.86s) - 248.357s = 12.503 seconds

    Benchmark Results - RTX 2080Ti - GTX 980Ti OC
    Total Rendering Time: 3 minutes 5.16 seconds
    CUDA device 1 (GeForce GTX 980 Ti):      514 iterations, 3.647s init, 177.934s render
    CUDA device 0 (GeForce RTX 2080 Ti):      1286 iterations, 3.307s init, 178.757s render
    Rendering Performance: 1800 / 178.757s = 10.070 iterations per second
    Loading Time: (180s + 5.16s) - 178.757s = 6.403 seconds

    Post edited by amadeus00 on
  • System Configuration
    System/Motherboard: Gigabyte Aorus Gaming 7
    CPU: AMD Ryzen Threadripper 1950X 16-Core Processor 3.40 GHz
    GPU: 4 X GV-N208TTURBO-11GC GeForce RTX2080Ti Turbo 11GB GDDR6 352Bit DisplayPort/HDMI/USB
    System Memory: Corsair Dominator Platinum 128GB (8x16GB) DDR4 3200MHz C16 Desktop Memory
    OS Drive: Samsung 970 EVO Plus Series - 1TB PCIe NVMe - M.2 Internal SSD (MZ-V7S1T0B/AM)
    Asset Drive: Samsung 970 EVO Plus Series - 1TB PCIe NVMe - M.2 Internal SSD (MZ-V7S1T0B/AM)
    Operating System: Windows 10 Build 1903
    Nvidia Drivers Version: 441.20
    Daz Studio Version: 4 12 0 86
    Optix Prime Acceleration: Off

    Benchmark Results - 4 X RTX 2080ti

    2019-11-14 22:17:28.943 Finished Rendering

    2019-11-14 22:17:28.993 Total Rendering Time: 1 minutes 14.95 seconds

    2019-11-14 22:17:36.520 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : Device statistics:

    2019-11-14 22:17:36.521 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 3 (GeForce RTX 2080 Ti):                450 iterations, 3.707s init, 67.418s render

    2019-11-14 22:17:36.521 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 0 (GeForce RTX 2080 Ti):                449 iterations, 3.680s init, 67.370s render

    2019-11-14 22:17:36.527 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 1 (GeForce RTX 2080 Ti):                453 iterations, 3.938s init, 66.954s render

    2019-11-14 22:17:36.527 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CUDA device 2 (GeForce RTX 2080 Ti):                448 iterations, 3.899s init, 66.291s render

    Iteration Rate: 26.86
    Loading Time: 7.95

    Benchmark Results - CPU Only

    2019-11-14 22:48:31.406 Finished Rendering

    2019-11-14 22:48:31.451 Total Rendering Time: 27 minutes 28.35 seconds

    2019-11-14 22:49:39.433 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : Device statistics:

    2019-11-14 22:49:39.440 Iray [INFO] - IRAY:RENDER ::   1.0   IRAY   rend info : CPU:            1800 iterations, 0.467s init, 1645.057s render

    Iteration Rate: 1.09
    Loading Time: 3.293 (I forgot to completely exit out of Daz)

Sign In or Register to comment.