Hi
I am running a Dual Xeon 3 Ghz Proc Quad core Mac Pro with 8GB Ram and Nvidia Quadro FX 4500 graphics option
Preview 2 scene 37:08 mins 3467165 Micro-triangles 800 x 600
Preview 3 all eight cores detected Min 1 max 16 threads set
Over 1hr 07 mins 13 sec 0 micro-triangles reported - same scene - But Image looks improved?? Sky renders better and surface detail is less grainy.
Will some scenes take longer to render? But no improvement using 8 cores.
Mac OSX 10.5.2 (Leopard)
If you need more info on the set-up let me know
Dean
Some scenes may take a little longer to render in the new version if it only uses 1 thread. But I would expect most scenes to render much more quickly if it is able to use more threads as it appears to on your machine.
I'd be interested to know how long it takes to render the same scene with different numbers for max threads. Tests with 0, 1 and 2 would be useful.
Matt
I'm experiencing similar problems with a dual 3Ghz Xeon Mac Pro OS X 10.5.2. 2Gb RAM
Render default scene (i.e no terrain) with 8 threads(cores) time 8 mins (TGTP 1.9.88.1)
Render default scene Time 1 min 34 secs (TGTP 1.9.04.1)
Report submitted to Planetside support but may have been caught in spam filter...??
Rgds
Chris
Chris, what about with max threads set to 0, 1 and 2? (Leave min threads at 1.)
Matt
Matt.
Will try 8 core @ work tomorrow, just ran similar test on old PPC dual core G5 ( Dual 2Ghz, Mac OS X 10.5.2) @ home similar results but not quite such a drastic differential.
TGTP 1.9.88.1 7 min 02 secs thread min set to 0
TGTP 1.9.88.1 7 min 04 secs thread min set to 1
TGTP 1.9.88.1 6 min 57 secs thread min set to 2
TGTP 1.9.04.1 5 min 55 secs
On Vista PC - Home Premium (Dell XPS 210, Intel Core 2 6300 1.8 Ghz, 4 Gb RAM ) Time to render 2 mins 27 secs ! (TGTP 1.9.88.1, 2 threads) 4 mins 27 secs (TGTP 1.9.88.1 1 thread)
As a side issue, even though I have purchased Deep+Animation, the version running under Vista still won't let me access the animation features. :'( (No great panic on that one, just a heads up)
Rgds
Chris
Doh .. you said max threads... too late tonight.. will repeat tomorrow !! ::)
I knew I shouldn't have added that min threads parameter... I had wanted to just have a max thread parameter to avoid these kinds of mixups ... :)
Quote from: red_planet on April 29, 2008, 07:34:26 PM
As a side issue, even though I have purchased Deep+Animation, the version running under Vista still won't let me access the animation features. :'( (No great panic on that one, just a heads up)
That's strange. Did you buy Deep first then later buy the Animation upgrade? The only thing I can think of is that you have two keys and it's using a different key to the one you think it's using.
Matt
Hi,
I've been doing some default scene renders on my dual G5 and Intel iMac. With two threads it is faster, but not dramatically so. With earlier development builds I saw more improvement, but they also weren't doing everything they should. I'm going to do some profiling to try and see what's happening.
One thing I have seen recently on an Apple developer performance mailing list is that there are some cases where the Mac OS is slower than Windows and Linux when it comes to certain multithreaded programming constructs. It's been a bit of a problem for a few people lately. These problems seem to be worse the more threads you're using, so this may be a case where an 8 core Mac runs into a bit of a dead end. However I just thought I would mention it, and I haven't yet been able to establish if these sorts of things are a problem for TG2 Mac. Unfortunately I don't have an 8 core Mac to test with :-).
Regards,
Jo
Hi Jo
I run Cinema4d using 8 cores on the Mac Pro and it is really fast and the machine is faster than my Armari x64 Windows machine also with dual 3 Ghz Quad cores and the same memory so I do not quite agree about the speed. I do know that some developers are finding optimisation difficult though.
I will try some tests later today with different threads - you say reduce the max threads?
Let me know any other tests you want me to try - I have not tried the terragen on the Windows 8 core - do you want me to try this?
Dean
A warning if you are using 8 threads. This alpha build isn't very well behaved with a large number of threads because it uses up to 150Mb per thread for the ray tracer (although it shouldn't be a problem with low resolution renders). 8 threads will therefore put an extra strain of 1 Gb above what would be required for 1 thread. The increased memory use may lead to extra virtual memory swapping and reduce performance (and increase the chance of running into memory allocation errors). I'm working on a way to share this memory between threads.
Matt
Maybe my usage of 4 to 6 in the Minimum Threads entry textbox is why I have been running into errors. Not sure. I did have a render come up blocky with the "render skip" error that was only using 1 Minimum Thread. If it helps, I could post the TGD here tonight (GMT -6).
Calico, do you have a system with 8 cores available? If not you're just needlessly increasing overhead. Currently even 4 cores eats quite a bit of RAM unfortunately. We're looking into enabling the switches on the development side that will make the /3GB switch work, as well as possibly allowing render tile threads to share data, but for now I'd recommend no more than 2 cores except for more simple scenes. There is also a memory leak that may be exacerbating the memory problems.
- Oshyan
@Oshyan, I have the Intel Core2 processor Q6600 (2.40Ghz 1066FSB) w/Quad Core Technology and 8MB cache. The memory is 4G of DDR2 SDRAM at 800MHz, which is setup with the /3GB switch.
With a quad core system you would never want to increase min threads over 4. There's just no good reason to. Currently you might even want to limit it to 2 for memory conservation. We're working on optimizing memory use and better utilization of quad core and higher systems should be possible in more memory-constrained situations in the future.
- Oshyan
Small addition...
I withheld further testing on the 8 core Mac Pro until such time as I had the opportunity to upgrade the RAM.
Now running with 6Gb.. no difference. Still 8 mins+ for a bare render with 8 threads. :( (The initial GI pass seems to be very quick ...it's the rendering that's seems to be taking all the time.)
but !!!
Bare render max 1 thread ....... 5 Min 20 secs
Bare render max 2 threads ..... 1 Min 52 secs !!!!
Bare render max 3 threads ..... 4 Min 10 secs
Bare render max 4 threads ..... 6 Min 35 secs
???
I'm sure you'll figure it out.. the guys at Blender did, although I'm reasonably sure that the underlying cause is most likely very different, it may be worth noting that it was when rendering with the internal renderer in Blender that the problem surfaced.
Rgds
Chris
Quote from: Oshyan on May 01, 2008, 11:09:23 PM
With a quad core system you would never want to increase min threads over 4. There's just no good reason to. Currently you might even want to limit it to 2 for memory conservation. We're working on optimizing memory use and better utilization of quad core and higher systems should be possible in more memory-constrained situations in the future.
- Oshyan
Hi Oshyan,
I have a quadcore and I always use min. 4 threads and I run into frozen renders (render screen runs white and program does not respond).
If I understand correctly lowering the min. thread to about 2 might solve this problem while still using all 4 cores?
Thanks in advance.
Martin
Yes Martin, lowering the minimum to 2 might indeed help, if it's due to memory issues (as it may very well be). The next build should give you more flexibility and control over memory use, as well as performing better in multi-core systems, not to mention the addition of "large address" support for those of us on x64 systems or using the /3GB switch.
- Oshyan
Quote from: Oshyan on May 21, 2008, 01:43:28 AM
Yes Martin, lowering the minimum to 2 might indeed help, if it's due to memory issues (as it may very well be). The next build should give you more flexibility and control over memory use, as well as performing better in multi-core systems, not to mention the addition of "large address" support for those of us on x64 systems or using the /3GB switch.
- Oshyan
Thanks Oshyan.
I forgot to mention, I was in a hurry to get from work :), that I have xp64, 8 GB RAM. So RAM shouldn't be that much of a problem :)
I'm currently rendering a very heavy scene with 7 VHQ, 3 HQ and 2 high-poly models @ 1200x800 with detail 1, AA 14 (grasses) and GI 2/2.
The render froze twice when using 1 instance with 4 min. threads and 16 max. threads, RAM usage was ~1.3 GB.
Yesterday I made a quick and dirty test to see the difference in adjusting the min. and max. threads.
I chose to render the default scene loaded by TG2.
- Using min. 1 and max. 2 results in 1 core of 4 rendering (25% CPU usage and ~110 MB RAM)
- Using min. 2 and max. 2 results in 3 cores, maybe 4? of 4 rendering (67% CPU usage and ~150 MB RAM)
- Using min. 2 and max. 4 results in 3 cores, maybe 4? of 4 rendering (72% CPU usage and ~190 MB RAM)
- Using min. 3 and max. 4 results in 4 cores of 4 rendering (100% CPU usage and ~220 MB RAM)
- Using min. 3 and max. 16 results in 4 cores of 4 rendering (10% CPU usage and ~260 MB RAM)
- Using min. 4 and max. 16 results in 4 cores of 4 rendering (100% CPU usage and >300 MB RAM)
The results aren't really predictable but it's obvious that even for a simple scene like this default scene the amount of RAM used increases quite much over increasing threads. Probably not to mention the difference when things are getting complicated.
Conclusively I decided to open 2 instances of TG with min. 2 and max. 2 threads and it is still rendering using ~1.3 GB RAM per instance :)
Oh...JOY!!! ;D
Quote from: Oshyan on May 21, 2008, 01:43:28 AM
...not to mention the addition of "large address" support for those of us on x64 systems or using the /3GB switch.
- Oshyan
Martin, your system will do quite well with the next update as it enables Large Address support, but until then 8GB of RAM only helps you much if you're running multiple instances.
- Oshyan
Quote from: red_planet on May 20, 2008, 10:24:34 AM
Small addition...
(The initial GI pass seems to be very quick ...it's the rendering that's seems to be taking all the time.)
Hi All,
this is what I'm getting when I use 8 threads; All 8 cores peg on the GI pass. (super fast) But when the render starts all 8 cores drop to about 20-30 percent usage per core. Core usage is fairly uniform across all 8 cores.
When I switch to 3 threads the render makes full use of 3 cores for a while but then starts to drop usage of the 3 cores.
It seems the more threads I add the less efficient the use of each core is.
It seems 2 threads is the sweet spot for now.
8 core MacPro 10 GB ram
base render 2 threads 1:36s
base render 3 threads 2:06s
base render 4 threads 2:57s
base render 8 threads 5:44s
I'm also a C4D user and Cinema uses all cores wide open.
Richard
Hi Richard/3space,
Thanks for the info.
Was this using a cloud layer? If so, would you be kind enough to repeat some of those tests with the cloud layer's acceleration cache set to "None (highest detail)"? (It's found on the Quality tab). We've recently identified the acceleration cache as a potential bottleneck for multi-threading.
Other things which might be useful to disable in separate tests are the Enviro Light or the ray-traced shadows option in the renderer.
Also, the size of the subdiv cache decreases per thread as the number of threads increases. The total size (shared among all threads) can be set in the render settings.
Early multi-threading tests on quad-core machines showed very good speedup factors over single-threaded renders, so I don't know if something has changed in recent builds or whether it is much more of a problem on some machines than others.
Matt
Hi Matt,
my first test where on a default scene without a cloud layer. I had the subdiv cach set to 1200 Mb.
__________________
__________________
Here are the results of my second test
base render + cloud layer "Optimal" 2 threads 2:20s
base render + cloud layer "highest" 2 threads 3:33s
base render + cloud layer "Optimal" 4 threads 4:09s
base render + cloud layer "highest" 4 threads 2:43s
base render + cloud layer "Optimal" 8 threads 4:33s
base render + cloud layer "highest" 8 threads 4:20s
__________________
__________________
Misc quick test
I noticed that the per core usage seemed more efficient when I added the clouds compared to my first renders that did not have a cloud layer; the usage of each core jump to around 40-60 percent when the cloud layer was added compared to a flat 30 percent with just the base terrain.
So I pointed the camera straight up and rendered just clouds and sky to see if the problem was just with the terrain. But just like with ground plane only, the core usage efficiency went down with just the cloud layer and stayed even throughout the rendering of the entire frame.
It seems when I include both clouds and terrain in the field of view the render uses more of the cpu resources. But if I just render the clouds or just the terrain all the cores flatline at about 30 precent each and stay at that point throughout the render.
Richard
Thanks Richard, this is useful information.
Matt
np, Let me know if you need more test. I would be glad to help.