AWS Thinkbox Discussion Forums

Krakatoa hardly taxing CPU

Hi All,

I am doing a long animation render, and I’ve noticed that although Krakatoa takes 2+ minutes per frame, it does not use 100% at all. I am on a 6-core (12 thread) system and it only seems to use 20%~ of the available power to render.

The scene has a lot of DOF and around 10 million particles. 4 different PRT loaders.

PRG: Rendering frame 2272
STS: Section “Collecting APM Volumes”:
STS: Total 00h 00m 00.000s Called 1 times Avg 00h 00m 00.000s
STS: Section “Evaluating Particle Objects”:
STS: Total 00h 00m 02.407s Called 1 times Avg 00h 00m 02.407s
STS: Section “Retrieving Particles”:
STS: Total 00h 00m 05.687s Called 1 times Avg 00h 00m 05.687s
PRG: Rendering 12323325 particles.
STS: Capping number of threads automatically at 12 because…
STS: Number of frame buffers per thread: 1
STS: Bytes per frame buffer:3145728
STS: Total Pysical Memory Detected: 34254987264
STS: Available memory for frame buffers: 18824081408
STS: Heuristic thread cap: 5984
STS: Effective user thread cap: 18446744073709551615
STS: Number of logical threads detected 12
STS: Section “Lighting:Updating Light”:
STS: Total 00h 00m 00.015s Called 1 times Avg 00h 00m 00.015s
STS: Section “Lighting:Sorting Particles”:
STS: Total 00h 00m 00.422s Called 1 times Avg 00h 00m 00.422s
STS: Section “Lighting:Calculating”:
STS: Total 00h 00m 01.172s Called 1 times Avg 00h 00m 01.172s
STS: Capping number of threads automatically at 12 because…
STS: Number of frame buffers per thread: 1
STS: Bytes per frame buffer:88473600
STS: Total Pysical Memory Detected: 34254987264
STS: Available memory for frame buffers: 17524813824
STS: Heuristic thread cap: 198
STS: Effective user thread cap: 18446744073709551615
STS: Number of logical threads detected 12
STS: Partition 0 starts at particle 0.
STS: Partition 1 starts at particle 562407.
STS: Partition 2 starts at particle 562441.
STS: Partition 3 starts at particle 562463.
STS: Partition 4 starts at particle 562473.
STS: Partition 5 starts at particle 562488.
STS: Partition 6 starts at particle 562506.
STS: Partition 7 starts at particle 562511.
STS: Partition 8 starts at particle 562528.
STS: Partition 9 starts at particle 562545.
STS: Partition 10 starts at particle 562583.
STS: Partition 11 starts at particle 562726.
STS: Section “Rendering:Matte”:
STS: Total 00h 00m 00.000s Called 0 times
STS: Section “Rendering:Sorting”:
STS: Total 00h 00m 02.891s Called 1 times Avg 00h 00m 02.891s
STS: Section “Rendering:Drawing”:
STS: Total 00h 02m 35.328s Called 1 times Avg 00h 02m 35.328s
PRG: Finished rendering frame: 2272
PRG: Rendering frame 2273
STS: Section “Collecting APM Volumes”:
STS: Total 00h 00m 00.000s Called 1 times Avg 00h 00m 00.000s
STS: Section “Evaluating Particle Objects”:
STS: Total 00h 00m 05.703s Called 1 times Avg 00h 00m 05.703s
STS: Section “Retrieving Particles”:
STS: Total 00h 00m 08.875s Called 1 times Avg 00h 00m 08.875s
PRG: Rendering 12346857 particles.
STS: Capping number of threads automatically at 12 because…
STS: Number of frame buffers per thread: 1
STS: Bytes per frame buffer:3145728
STS: Total Pysical Memory Detected: 34254987264
STS: Available memory for frame buffers: 18762565632
STS: Heuristic thread cap: 5964
STS: Effective user thread cap: 18446744073709551615
STS: Number of logical threads detected 12
STS: Section “Lighting:Updating Light”:
STS: Total 00h 00m 00.000s Called 1 times Avg 00h 00m 00.000s
STS: Section “Lighting:Sorting Particles”:
STS: Total 00h 00m 00.454s Called 1 times Avg 00h 00m 00.454s
STS: Section “Lighting:Calculating”:
STS: Total 00h 00m 01.250s Called 1 times Avg 00h 00m 01.250s
STS: Capping number of threads automatically at 12 because…
STS: Number of frame buffers per thread: 1
STS: Bytes per frame buffer:88473600
STS: Total Pysical Memory Detected: 34254987264
STS: Available memory for frame buffers: 17298368512
STS: Heuristic thread cap: 195
STS: Effective user thread cap: 18446744073709551615
STS: Number of logical threads detected 12
STS: Partition 0 starts at particle 0.
STS: Partition 1 starts at particle 561828.
STS: Partition 2 starts at particle 561921.
STS: Partition 3 starts at particle 561926.
STS: Partition 4 starts at particle 561931.
STS: Partition 5 starts at particle 561933.
STS: Partition 6 starts at particle 561935.
STS: Partition 7 starts at particle 561952.
STS: Partition 8 starts at particle 561964.
STS: Partition 9 starts at particle 561992.
STS: Partition 10 starts at particle 562040.
STS: Partition 11 starts at particle 562206.
STS: Section “Rendering:Matte”:
STS: Total 00h 00m 00.000s Called 0 times
STS: Section “Rendering:Sorting”:
STS: Total 00h 00m 02.860s Called 1 times Avg 00h 00m 02.860s
STS: Section “Rendering:Drawing”:
STS: Total 00h 02m 35.046s Called 1 times Avg 00h 02m 35.046s
PRG: Finished rendering frame: 2273
PRG: Rendering frame 2274
STS: Section “Collecting APM Volumes”:
STS: Total 00h 00m 00.000s Called 1 times Avg 00h 00m 00.000s
STS: Section “Evaluating Particle Objects”:
STS: Total 00h 00m 02.687s Called 1 times Avg 00h 00m 02.687s
STS: Section “Retrieving Particles”:
STS: Total 00h 00m 09.484s Called 1 times Avg 00h 00m 09.484s
PRG: Rendering 12394711 particles.
STS: Capping number of threads automatically at 12 because…
STS: Number of frame buffers per thread: 1
STS: Bytes per frame buffer:3145728
STS: Total Pysical Memory Detected: 34254987264
STS: Available memory for frame buffers: 17219801088
STS: Heuristic thread cap: 5474
STS: Effective user thread cap: 18446744073709551615
STS: Number of logical threads detected 12
STS: Section “Lighting:Updating Light”:
STS: Total 00h 00m 00.000s Called 1 times Avg 00h 00m 00.000s
STS: Section “Lighting:Sorting Particles”:
STS: Total 00h 00m 00.422s Called 1 times Avg 00h 00m 00.422s
STS: Section “Lighting:Calculating”:
STS: Total 00h 00m 01.187s Called 1 times Avg 00h 00m 01.187s
STS: Capping number of threads automatically at 12 because…
STS: Number of frame buffers per thread: 1
STS: Bytes per frame buffer:88473600
STS: Total Pysical Memory Detected: 34254987264
STS: Available memory for frame buffers: 15985299456
STS: Heuristic thread cap: 180
STS: Effective user thread cap: 18446744073709551615
STS: Number of logical threads detected 12
STS: Partition 0 starts at particle 0.
STS: Partition 1 starts at particle 561528.
STS: Partition 2 starts at particle 561575.
STS: Partition 3 starts at particle 561585.
STS: Partition 4 starts at particle 561602.
STS: Partition 5 starts at particle 561613.
STS: Partition 6 starts at particle 561621.
STS: Partition 7 starts at particle 561640.
STS: Partition 8 starts at particle 561653.
STS: Partition 9 starts at particle 561658.
STS: Partition 10 starts at particle 561669.
STS: Partition 11 starts at particle 561732.
STS: Section “Rendering:Matte”:
STS: Total 00h 00m 00.000s Called 0 times
STS: Section “Rendering:Sorting”:
STS: Total 00h 00m 03.000s Called 1 times Avg 00h 00m 03.000s
STS: Section “Rendering:Drawing”:
STS: Total 00h 02m 35.485s Called 1 times Avg 00h 02m 35.485s
PRG: Finished rendering frame: 2274
PRG: Rendering frame 2275
STS: Section “Collecting APM Volumes”:
STS: Total 00h 00m 00.000s Called 1 times Avg 00h 00m 00.000s
STS: Section “Evaluating Particle Objects”:
STS: Total 00h 00m 06.422s Called 1 times Avg 00h 00m 06.422s
STS: Section “Renderer::Open()”:
STS: Total 00h 00m 00.000s Called 1 times Avg 00h 00m 00.000s
STS: Section “Renderer::Close()”:
STS: Total 00h 00m 01.922s Called 1 times Avg 00h 00m 01.922s
STS: Section “Renderer::Render()”:
STS: Total 00h 08m 42.561s Called 4 times Avg 00h 02m 10.640s

How about disk / network usage? The ideal bottleneck to getting work done is obviously the CPU or graphics card, but there are other places that could slow you down. We’ve actually knocked networks over loading particle sets, so I’d start checking that network and file server usage.

Hi lostpixels.
Is the 20% CPU usage you’re seeing consistent throughout the rendering of a single frame, or does it vary significantly over the course of the render?
Also, if you open Resource Monitor and watch the CPU tab while rendering, does the load on the CPU appear to be balanced between the different cores, or is it much higher on one or a few?

So the files are local, and the disk usage is not a factor according to the task manager. Attached is pic of that, with the per-core readings from the resource monitor too. It just seems like every core is operating, but at around 50%~ of what it could be doing.

For that total CPU usage chart, do you know approximately when the render started and ended in it?

Privacy | Site terms | Cookie preferences