Distributed Arras/Moonray on Xeon Phi cluster? #25
Closed
ryanrussell
started this conversation in
Ideas
Replies: 1 comment 2 replies
-
|
Sadly, the work on xeon phi has been stagnant for years. While most of the machines at DreamWorks support AVX512 within the CPU architecture (independent of xeon phi), we have not tried even that in quite some time. MoonRay does not make run-time SIMD decisions (yet), and we compile explicitly for AVX2 support. I would be very interested to see if somebody can build MoonRay with AVX512 support. For clarification, Embree does dynamic SIMD decisions, and ISPC allows compilation with multiple SIMD architectures. We do not have that ISPC feature enabled. |
Beta Was this translation helpful? Give feedback.
2 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
While skimming code, I couldn't help but notice some references to xeon phi.
Not sure if the current codebase could be packaged to run on thos MICs, but this is a really interesting application if so...
Three questions:
Is anyone from the engineering team from dreamworks able to share whether the codebase can still be run on xeon phi's?
(If yes), are there any kind of benchmarks(real question: are they so slow that it's not worth the electricity?)
(If yes), Are there any guides that could be cobbled together around this?
Would love to chat if anyone has devops knowledge around this :)
Beta Was this translation helpful? Give feedback.
All reactions