Further optimizations for ARM M1 Architecture
I am testing SQX on Apple's M1 architecture. So far everything works. However, it is still not optimized for ARM as the Activity Monitor is showing that the current architecture is for Intel processors. Im sure performance will increase after optimizing for the newer architecture. Im assuming the current build is running through Rosetta which has not been great for sustained workloads.
Also, it would be great if you could try and see if the additional cores from the GPU and Neural Engine can be used. I am not sure if the same limitations are present because the additional cores are all on the same chip.
Attachments
No attachments
-
Votes +2
-
Project StrategyQuant X
-
Type Feature
-
Status New
-
Priority Normal
History
KL
JH
KL
kainc301
22.03.2022 22:45
With the new M1 Ultra chip and the next Mac Pro on the horizon, this is going to become much more important soon.
Mark, do you think that the unified memory in the new architecture could allow for inclusion of the GPU cores? Correct me if I'm wrong but the previous limitation had to do with normal GPUs using a separate memory cache.
E
Votes: +2
© Copyright. All rights reserved. ProjectPanel.com
