Optimal Professional Render Worksation

[bill_lake]

I want to get 100% CPU utilization during AS renders, at any cost.
I would like this thread to be a very knowledgeable, technical discussion oriented to those willing to build a high-end workstation optimized exclusively for Anime Studio Pro.


I am looking for a very technical explanation of if/how the AS rendering engine makes use of multiple processors, multiple cores, hyper-threading, cuda, windows resources, how it manages reads/writes, etc.

Why:
I have successfully optimized my workstation for professional video editing (PPr and AE) in order to gain 100% CPU utilization. When I render in AS, I only get 66% CPU and 30% RAM utilization and I cannot see any I/O bottlenecks and I cannot see any cores (8) parked.

I want to achieve 100% CPU utilization to reduce render time substantially and speed my workflow and I a plan to build a workstation optimized for AS rendering. Before I build it, I want to understand what AS needs. In theory, the CPU should run near 100% if it is not waiting to get or write data and I cannot see an I/O bottlenecks.

My Use:
I typically create 10-15, 60-second animation clips per day, rendering to .png sequences. I could produce about 50% more if I weren't waiting for the renders

My current workstation:
i7 930 2.8Ghz 24GB RAM nVidia 480GTX
internal drives:
C: RAID-0 2x500MB 7200rpm : OS/Apps
B: SSD (evo 840 128GB): Anime Studio Pro 10.0 app
D: RAID-0 2x1TB 10000rpm: Project Files, images, audio.
External
F: esata raid0 2x1TB 7200rpm: renders (.png sequences, or uncompressed .avi)
(no drives are used more than 30% except C: which is 60%)
I have no apps running in the background, no virus protection, - running as clean as I can. This is a dedicated video/animation workstation.

Things I have tried:
After seeing a lot of C: drive activity during renders (when rendering to a different drive), I moved the AS app to B: (SSD) and that seemed to help improve processor utilization (up to 66% from around 50%). I've tried rendering to the SSD with the app on a different drives - little impact. I still notice a high amount of C: activity during rendering, even though the app is on B:. Still - I do not see I/O bottlenecks. I suspect the render engine doesn't use the cores well or it has timing or swapping issues referencing libraries. I know that I could move to faster drives, faster processors, etc. but I want to do this objectively so that I gain 100% processor utilization.

Has anyone researched this? Is there a SmithMicro tech that monitors this site? Does SmithMicro help pros who's businesses have committed to the product? Adobe provides detailed information about how to optimize for their rendering engine(s) so that professionals can invest wisely - I wish SmithMicro would do the same, but everything I find is purely consumer-lite info.

Thanks,

-Bill

[toast]

I would also like to know how to do this. I have an $8,000 machine built just for rendering in maya and only get 20% CPU. Renders take hours and it drives me up a wall! I'm rendering about 90 sec of heavy duty stuff. It would be great if SM listed what options we had!

[SimplSam]

I would also like to know how to do this. I have an $8,000 machine built just for rendering in maya and only get 20% CPU.

One option would be to swap your $8K machine with my old $80 machine which will quite happily run @ 100% CPU!

[funksmaname]

I would also like to know how to do this. I have an $8,000 machine built just for rendering in maya and only get 20% CPU.

One option would be to swap your $8K machine with my old $80 machine which will quite happily run @ 100% CPU!

brilliant.

[toast]

I would also like to know how to do this. I have an $8,000 machine built just for rendering in maya and only get 20% CPU.

One option would be to swap your $8K machine with my old $80 machine which will quite happily run @ 100% CPU!

That was funny!


But really, this is extremely frustrating when using in a professional environment.