Graphicscard and render... Again...
GroveFast
Posts: 19
Yes again... I know there are multible posts on this, but I actually couln't find any recent good posts on this.
I want to end up with the ability to have 3 HD characters (cloth and hair) a decent sized set and a HDRI background + 3-4 lights. My PC is: i7 CPU, GTX970 4GB and 16GB RAM. As I understnad form reading posts, that 4GB VRAM is not quite enough for what I want to do.
My problem is that I seem to hit the limit of my graphicscard VRAM with only very few elements in the scene. If i put a basic HD character with clothes and hair in a scene, my current 4GB is already uses up and there is a partiall fallback to CPU render. For larger scenes (just adding one more character and a backgorund) the GPU is not used at all.
I'm thinking about buying a new card (GTX2070 with 8GB), but with the rate Daz is eating VRAM, I'm not sure even that card can handle what I wnat to achive (3 HD characters (cloth and hair) a decent sized set and a HDRI background + 3-4 lights)
Either there is something I do not understand or there is a setting somewhere I need to have a look at.
Let me show some logs and utilization. Scene consist of Noemi HD character, Summer sidetail hair and some random G8 clothes
I do 5 Runs as such:
RUN 1 (Lading basic figure - render)
RUN 2 (cancel render - restart render)
RUN 3 (Save file - close/reopen Daz - load file - render)
RUN 4 (restart PC - restart Daz - load file - render)
RUN 5 (stop render - switch to GPU only - render)
#################################
RUN 1 (Lading basic figure - render)
line 1, render starts at line 1183
#################################2020-04-22 09:03:42.218 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : CUDA device 0 (GeForce GTX 970): Scene processed in 16.834s
2020-04-22 09:03:42.218 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - IRAY:RENDER :: 1.5 IRAY rend error: CUDA device 0 (GeForce GTX 970): out of memory (while allocating memory)
2020-04-22 09:03:42.218 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - IRAY:RENDER :: 1.5 IRAY rend error: CUDA device 0 (GeForce GTX 970): Failed to allocate 915.18 MiB
2020-04-22 09:03:42.234 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 158.203 MiB for frame buffer
2020-04-22 09:03:42.421 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 681.961 MiB of work space (808k active samples in 0.173s)
2020-04-22 09:03:42.453 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : CUDA device 0 (GeForce GTX 970): Used for display, optimizing for interactive usage (performance could be sacrificed)
2020-04-22 09:03:46.139 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : Allocating 1-layer frame buffer
2020-04-22 09:03:46.358 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - POST:RENDER :: 1.0 POST rend error: A CUDA error occurred: out of memory
2020-04-22 09:03:46.358 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [WARNING] - POST:RENDER :: 1.0 POST rend warn : Failed to allocate device memory, falling back to CPU
2020-04-22 09:03:46.358 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 134.789 MiB of work space (159k active samples in 0.001s)
2020-04-22 09:10:08.327 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00304 iterations after 402.948s.
2020-04-22 09:10:21.705 Saved image: C:\Users\justa\AppData\Roaming\DAZ 3D\Studio4\temp\render\r.png
2020-04-22 09:10:21.705 Finished Rendering
2020-04-22 09:10:21.752 Total Rendering Time: 7 minutes 7.78 seconds
2020-04-22 09:10:26.502 WARNING: ..\..\..\..\..\src\sdksource\cloud\dzcloudtasknotifier.cpp(178): peer performed orderly shutdown errno=0
2020-04-22 09:10:37.088 Saved image: C:\Users\justa\Documents\DAZ 3D\Studio\Render Library\test1.png
2020-04-22 09:10:37.322 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Device statistics:
2020-04-22 09:10:37.338 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : CUDA device 0 (GeForce GTX 970): 247 iterations, 17.460s init, 397.093s render
2020-04-22 09:10:37.338 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : CPU: 57 iterations, 15.705s init, 398.834s render
## 304 iterations in 402 secs
##################################################################
RUN 2 (cancel render - restart render)
line 1411
##################################################################
2020-04-22 09:12:47.937 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : CUDA device 0 (GeForce GTX 970): Scene processed in 15.497s
2020-04-22 09:12:47.937 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - IRAY:RENDER :: 1.5 IRAY rend error: CUDA device 0 (GeForce GTX 970): out of memory (while allocating memory)
2020-04-22 09:12:47.937 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - IRAY:RENDER :: 1.5 IRAY rend error: CUDA device 0 (GeForce GTX 970): Failed to allocate 915.18 MiB
2020-04-22 09:12:47.953 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 158.203 MiB for frame buffer
2020-04-22 09:12:48.031 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 332.121 MiB of work space (393k active samples in 0.011s)
2020-04-22 09:12:48.031 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : CUDA device 0 (GeForce GTX 970): Used for display, optimizing for interactive usage (performance could be sacrificed)
2020-04-22 09:12:49.892 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : Allocating 1-layer frame buffer
2020-04-22 09:12:50.127 Iray [INFO] - IRAY:RENDER :: 1.5 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 221.484 MiB of work space (262k active samples in 0.087s)
2020-04-22 09:12:50.236 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00001 iterations after 17.808s.
2020-04-22 09:12:51.970 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00002 iterations after 19.545s.
2020-04-22 09:18:41.222 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00300 iterations after 368.793s.
2020-04-22 09:19:01.527 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00305 iterations after 389.098s.
2020-04-22 09:19:04.651 Saved image: C:\Users\justa\AppData\Roaming\DAZ 3D\Studio4\temp\render\r.png
2020-04-22 09:19:04.651 Finished Rendering
2020-04-22 09:19:04.698 Total Rendering Time: 6 minutes 41.24 seconds
2020-04-22 09:19:08.181 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Device statistics:
2020-04-22 09:19:08.181 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : CUDA device 0 (GeForce GTX 970): 250 iterations, 16.603s init, 373.631s render
2020-04-22 09:19:08.197 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : CPU: 55 iterations, 14.850s init, 375.368s render
## 305 iterations in 389 secs##################################################################
RUN 3 (Save file - close/reopen Daz - load file - render)
line 1640
##################################################################2020-04-22 09:24:41.482 Iray [INFO] - IRAY:RENDER :: 1.6 IRAY rend info : CUDA device 0 (GeForce GTX 970): Scene processed in 16.195s
2020-04-22 09:24:41.482 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - IRAY:RENDER :: 1.6 IRAY rend error: CUDA device 0 (GeForce GTX 970): out of memory (while allocating memory)
2020-04-22 09:24:41.482 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - IRAY:RENDER :: 1.6 IRAY rend error: CUDA device 0 (GeForce GTX 970): Failed to allocate 915.18 MiB
2020-04-22 09:24:41.497 Iray [INFO] - IRAY:RENDER :: 1.6 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 158.203 MiB for frame buffer
2020-04-22 09:24:41.622 Iray [INFO] - IRAY:RENDER :: 1.6 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 628.488 MiB of work space (744k active samples in 0.114s)
2020-04-22 09:24:41.669 Iray [INFO] - IRAY:RENDER :: 1.6 IRAY rend info : CUDA device 0 (GeForce GTX 970): Used for display, optimizing for interactive usage (performance could be sacrificed)
2020-04-22 09:24:45.325 Iray [INFO] - IRAY:RENDER :: 1.6 IRAY rend info : Allocating 1-layer frame buffer
2020-04-22 09:24:45.530 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - POST:RENDER :: 1.0 POST rend error: A CUDA error occurred: out of memory
2020-04-22 09:24:45.530 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [WARNING] - POST:RENDER :: 1.0 POST rend warn : Failed to allocate device memory, falling back to CPU
2020-04-22 09:24:45.530 Iray [INFO] - IRAY:RENDER :: 1.6 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 186.258 MiB of work space (220k active samples in 0.004s)
2020-04-22 09:24:45.780 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00001 iterations after 20.486s.
2020-04-22 09:30:26.278 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00294 iterations after 360.984s.
2020-04-22 09:30:41.723 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00300 iterations after 376.444s.
2020-04-22 09:30:48.643 Saved image: C:\Users\justa\AppData\Roaming\DAZ 3D\Studio4\temp\render\r.png
2020-04-22 09:30:48.643 Finished Rendering
2020-04-22 09:30:48.690 Total Rendering Time: 6 minutes 34.88 seconds
2020-04-22 09:30:49.143 WARNING: ..\..\..\..\..\src\sdksource\cloud\dzcloudtasknotifier.cpp(178): peer performed orderly shutdown errno=0
2020-04-22 09:30:57.080 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Device statistics:
2020-04-22 09:30:57.080 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : CUDA device 0 (GeForce GTX 970): 235 iterations, 17.103s init, 364.514s render
2020-04-22 09:30:57.080 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : CPU: 65 iterations, 15.375s init, 366.228s render
## 300 iterations in 376 secs##################################################################
RUN 4 (restart PC - restart Daz - load file - render)
line 4163
##################################################################
2020-04-22 09:38:16.393 Iray [INFO] - IRAY:RENDER :: 1.2 IRAY rend info : CUDA device 0 (GeForce GTX 970): Scene processed in 15.809s
2020-04-22 09:38:16.393 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - IRAY:RENDER :: 1.2 IRAY rend error: CUDA device 0 (GeForce GTX 970): out of memory (while allocating memory)
2020-04-22 09:38:16.408 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - IRAY:RENDER :: 1.2 IRAY rend error: CUDA device 0 (GeForce GTX 970): Failed to allocate 915.18 MiB
2020-04-22 09:38:16.424 Iray [INFO] - IRAY:RENDER :: 1.2 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 158.203 MiB for frame buffer
2020-04-22 09:38:16.580 Iray [INFO] - IRAY:RENDER :: 1.2 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 709.91 MiB of work space (841k active samples in 0.153s)
2020-04-22 09:38:16.627 Iray [INFO] - IRAY:RENDER :: 1.2 IRAY rend info : CUDA device 0 (GeForce GTX 970): Used for display, optimizing for interactive usage (performance could be sacrificed)
2020-04-22 09:38:19.941 Iray [INFO] - IRAY:RENDER :: 1.2 IRAY rend info : Allocating 1-layer frame buffer
2020-04-22 09:38:20.175 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - POST:RENDER :: 1.0 POST rend error: A CUDA error occurred: out of memory
2020-04-22 09:38:20.175 Iray [INFO] - IRAY:RENDER :: 1.2 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 140.273 MiB of work space (166k active samples in 0.018s)
2020-04-22 09:38:20.191 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [WARNING] - POST:RENDER :: 1.0 POST rend warn : Failed to allocate device memory, falling back to CPU
2020-04-22 09:38:20.456 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00001 iterations after 19.872s.
2020-04-22 09:38:21.472 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00002 iterations after 20.881s.
2020-04-22 09:46:47.134 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00321 iterations after 526.550s.
2020-04-22 09:46:50.133 Saved image: C:\Users\justa\AppData\Roaming\DAZ 3D\Studio4\temp\render\r.png
2020-04-22 09:46:50.133 Finished Rendering
2020-04-22 09:46:50.180 Total Rendering Time: 9 minutes 1.8 seconds
2020-04-22 09:49:07.362 WARNING: ..\..\..\..\..\src\sdksource\cloud\dzcloudtasknotifier.cpp(178): peer performed orderly shutdown errno=0
2020-04-22 09:49:19.395 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Device statistics:
2020-04-22 09:49:19.395 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : CUDA device 0 (GeForce GTX 970): 237 iterations, 16.560s init, 511.340s render
2020-04-22 09:49:19.395 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : CPU: 84 iterations, 15.288s init, 512.391s render
## 321 iterations in 526 secs##################################################################
RUN 5 (stop render - switch to GPU only - render)
line 6354
##################################################################
2020-04-22 10:02:41.692 Iray [INFO] - IRAY:RENDER :: 1.4 IRAY rend info : CUDA device 0 (GeForce GTX 970): Scene processed in 14.498s
2020-04-22 10:02:41.692 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - IRAY:RENDER :: 1.4 IRAY rend error: CUDA device 0 (GeForce GTX 970): out of memory (while allocating memory)
2020-04-22 10:02:41.692 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - IRAY:RENDER :: 1.4 IRAY rend error: CUDA device 0 (GeForce GTX 970): Failed to allocate 915.18 MiB
2020-04-22 10:02:41.708 Iray [INFO] - IRAY:RENDER :: 1.4 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 158.203 MiB for frame buffer
2020-04-22 10:02:41.755 Iray [INFO] - IRAY:RENDER :: 1.4 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 712.441 MiB of work space (844k active samples in 0.048s)
2020-04-22 10:02:41.755 Iray [INFO] - IRAY:RENDER :: 1.4 IRAY rend info : CUDA device 0 (GeForce GTX 970): Used for display, optimizing for interactive usage (performance could be sacrificed)
2020-04-22 10:02:44.988 Iray [INFO] - IRAY:RENDER :: 1.4 IRAY rend info : Allocating 1-layer frame buffer
2020-04-22 10:02:45.144 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - POST:RENDER :: 1.0 POST rend error: A CUDA error occurred: out of memory
2020-04-22 10:02:45.144 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [WARNING] - POST:RENDER :: 1.0 POST rend warn : Failed to allocate device memory, falling back to CPU
2020-04-22 10:02:45.144 Iray [INFO] - IRAY:RENDER :: 1.4 IRAY rend info : CUDA device 0 (GeForce GTX 970): Allocated 140.801 MiB of work space (166k active samples in 0.001s)
2020-04-22 10:02:45.347 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00001 iterations after 18.160s.
2020-04-22 10:02:47.316 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00002 iterations after 20.125s....
2020-04-22 10:10:49.928 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00234 iterations after 502.736s.
2020-04-22 10:11:39.698 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00258 iterations after 552.513s.
2020-04-22 10:12:34.137 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00284 iterations after 606.943s.
2020-04-22 10:13:14.245 WARNING: ..\..\..\..\..\src\sdksource\cloud\dzcloudtasknotifier.cpp(178): peer performed orderly shutdown errno=0
2020-04-22 10:13:34.758 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Received update to 00313 iterations after 667.561s.
2020-04-22 10:13:42.991 Saved image: C:\Users\justa\AppData\Roaming\DAZ 3D\Studio4\temp\render\r.png
2020-04-22 10:13:42.991 Finished Rendering
2020-04-22 10:13:43.025 Total Rendering Time: 11 minutes 27.45 seconds
2020-04-22 10:14:56.442 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : Device statistics:
2020-04-22 10:14:56.442 Iray [INFO] - IRAY:RENDER :: 1.0 IRAY rend info : CUDA device 0 (GeForce GTX 970): 313 iterations, 14.841s init, 659.271s render## 313 iterations in 667 secs
Attached is the complete log
I'm specially worried about this message:
2020-04-22 10:02:45.144 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [ERROR] - POST:RENDER :: 1.0 POST rend error: A CUDA error occurred: out of memory
2020-04-22 10:02:45.144 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(305): Iray [WARNING] - POST:RENDER :: 1.0 POST rend warn : Failed to allocate device memory, falling back to CPU
And that during rendering I see using OpenHardwareMonitor, that my GPU memory is 90-98% loaded. See attached OpenHardwareMonitor printout of the 5 runs.
Again I'm worried that buing a 2070 with 8GB VRAM, with the demand I have, will agian just fallback to CPU.
Daz 3D is part of 
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
First there is no such thing as "partial fallback" Either the VRAM is exceeded and the GPU isn't used at all or it isn't.
8Gb is going to be tight for 3 HD characters and an environment. But why do you need 3 HD characters in the same scene? The HD morphs aren't visible except in closeups. There's no way to have 3 characters in a closeup.
You'll either need to go for a 2080ti or a RTX Titan, or a Quadro or wait till the beta comes out with NVLink VRAM pooling and 2 2070 Supers' will get you around 14Gb.
HD is noticable at mid range, not just close-up (if you're looking). Anyway 3, 4, 5 characters are doable in 8Gb without falling back. A 2070 would be OK.
Thanks for the answers. Sonuds like I need to get one of those 11GB cards then :) Seriously, this is just a hobby, and I cant get a 2080ti anytime soon. The 2070 seems like a nice tradoff in terms of price, so I need to start learning the art og limitations. Starting with my 4GB card seems like a good idea, then when I get the 8GB 2070 it will feel like a huge boost.
Some considerations:
1. I need 2 graphics cards. one for the screen and one for rendering. My 1440p screen takes almost 500MB of VRAM, so moving that onto another card would seem like a good thing. I will therefor use a GTX 950 I have lying around for the screen and the GTX970 only for render and also add the 2070 when I get that. That means when rendring small scens (below 4GB) I have 2 cards for rendering :) :)
2. Noemi characters Freckles textures are abolutly HUGE. Removing them is way better on the VRAM usage. So I guess again the art of limitations - only using HD and large textures when absolutly needed.
3. I think props and characters not even visible but still in the scene, are still calculated. So removing anything NOT visible (unless they interact with the light in a significant way) would be good as well (Am I correct in this assumption?)
4. Regarding the fall back, this is a good point. I am confused about the messages in the log:
In this instance the GC is stille rendering despite the cryptic message "out of memory (while allocating memory)" and "Failed to allocate 915.18 MiB"
In the next instance the GC is NOT rendering, and the difference would be this message: "rend error: Unable to allocate 67108864 bytes from 223256357 bytes of available device memory"
So in both cases the GC is apparently out of memory but despite that, it will still render in the first instance but not the second. I also see that despite adding or removing things in the scene, the VRAM momory consimption is still close to 100% - this is why my conlusion was that some king of partial rendring is happening.
I think that I must conclude that the render details in the log file should not always be taken literarly and that experience above all is better than trying to analyse the log file to anticipate what the outcome is.
I will try to gather some more experience and make it a target to stay below the 4GB VRAM limit - then when I get the 2070 Daz render should feel like a breeze
Hopefully other forum members will find this usefull
In order to get a high resolution for your DAZ Images you need a SSD, Liquid Cooling system, Core i9, Ram 64 Gig, GForce 2080 video Card. With a system like this, your renders would only take about 1 minute 30 seconds.
I got a 2080 super, fastest render times I get is like 15 minutes, that's using HDRI instead of a geometry scene, and only one girl lol. Not sure what kind of render is only going to take a min and a half, maybe at like a tiny resolution or something.
It is always going to be a close thing on a 4Gb card. For iRay that just isn't much at all.
Also you do not a SSD, water cooled components, high end CPU's, more than 16Gb of RAM or a specific Nvidia based card to get HD renders. By and large the rendering process barely uses any system resources except the GPU and VRAM. Once you're using an 8Gb card the rest is just about how long you're willing to wait.
Interior rendering using Architectural ON makes crying all your expensive cards...
For HD Genesis8 there are tricks when Vram is limited, one is resize textures, other is eliminate bump or displacement maps, other is using base geometry and not high resolution.
here, many figures, 2 are G8, and 2 are G1, reducing textures saves VRAM:
PC specs, nothing new, really obsolete crap machine
IMO, the minimum reccommended should be 8GB; note that I'm saying reccommended.
So, just want to toss this out there, I am running into similar issues with failing to allocate memory that I am trying to figure out, and I am using an RTX 2080Ti with 15Gb available VRAM. When things are working properly, having a small amount of VRAM (say, 8GB) should not prevent you from being able to render scenes, it should just cause them to take longer because of all of the juggling the render has to do moving things into and out of VRAM. When Iray runs properly, it should, for a complex scene, be using close to 99% of the VRAM with no issue.
The 2080Ti has 11GB - I think you are confusing system RAM and GPU RAM
I'm not sure I fully understand what you mean in the highlighted text. Just to clarify any possible future confusion, Iray needs to load the entire scene into the GPU (video card) memory. There is no swapping of scene data in/out of GPU and system memory during the render. Iray (in GPU only mode) does reserve/use one CPU core during the rendering process, but this isn't to move any scene data back and forth between the GPU and system memory. IIRC, the CPU is getting information from the GPU to write the image to the image file on the system (hard drive), and may also be being using it to calculate convergence. When running in GPU and CPU mode (using both to render the scene), GPU rendering is done the same way as in GPU only mode, but the CPU is also used to render the image (with a separate copy of the scene held in system memory for CPU use). Image/render data from the GPU and CPU are combined to make the rendering image which is used to calculate convergence and create the final image.
Iray in GPU mode is not like other some other render engines (i.e. Blender Cycles and Octane Render) that allow the use of system RAM to store and swap textures and geometry that can't fit into GPU memory. It's also important to note that even in GPU only mode, a copy of the compiled scene for use by Iray (the data that is sent to the GPU) is held in system RAM until the render is completed/closed. This is why you need extra system RAM (at least 2x or more than what your GPU has) to render efficiently with Iray. For example, if you have a 2080ti (11Gb memory), to ensure you don't use swap disk which severely slows down the entire system, and can cause instability (where you run out of physical system RAM, and Windows swaps System memory to your hard drive), you would want a system with a minimum of 24 Gb of system RAM.
Hope this clears thing up a little bit, and I didn't just make things more complicated.
To expand it's one thread not one core. While it is doing some things it's also needed just to keep iRay running. software cannot run just on the GPU. there has to be a process running on the cPU or the OS will clear out the GPU.
Question I will have budget to upgrade to 2080TI from my 1080. How much would this realistically increase render speed and do my specs like motherboard and power supply meet the requirements?
The 2080ti has more and newer CUDA so it will render faster. It also has the RTX features. So how much it will improve your render times depends greatly on what scenes you render. Ray tracing is much more complicated the more stuff light can reflect off in a scene. so indoor scenes, with walls, ceilings etc., tend to be accelerated much more than outdoor ones.
What about my hardware will it meet the requirement? I find full indoor scenes can be painfully long. I know HRDI or using a photo in the BG plain go much faster
If 16Gb of RAM has worked with your 1080ti it will work just as well with the 2080ti. The PSU will work fine as well, it won't if you keep the 1080ti and try to use both. For that you'd need a higher output PSU.
I have 32G in my tower. Thanks selling my 1080TI is part of why I can budget the purchase. It would just over 1/3 plus my wife's pandemic payments have been good to us;) I work in healthcare so lots of OT! I also have a 1070 laptop but just use it for scene building..
i did THIS with my 2070. now context. I use Scene Optimizer and hid alot of stuff that clearly wasn't going to be in the render. threadripper 1920x 32gb ram
I may be wrong but I believe Win10 is still hoging vram.