-
Notifications
You must be signed in to change notification settings - Fork 28
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Rework the Scheduling Example #468
Conversation
Codecov Report
❗ Your organization is not using the GitHub App Integration. As a result you may experience degraded service beginning May 15th. Please install the Github App Integration for your organization. Read more. @@ Coverage Diff @@
## master #468 +/- ##
==========================================
- Coverage 86.42% 86.37% -0.06%
==========================================
Files 78 78
Lines 17929 17933 +4
==========================================
- Hits 15496 15490 -6
- Misses 2433 2443 +10
|
First off, thanks so much for rewriting this. I have a tendency to just plow through with a ton of nitpicky criticisms, so please don't take it as anything other than constructive. This is so much better than what we had before. I'm reading through the rendered version alone for comments here...
Btw, for what K and machine specs are you reporting those perf numbers? It might be a good idea to make sure we're not doing something wrong performance-wise. |
Dumb thing this made me notice (but was there all along): is it really only a 2.2x speedup over the trivial triple for loop? |
Okay, addressed the comments! As far as the number go, it seems that Rosetta 2 cannot emulate AVX2 instructions so I can get rid of the output from running that command for now and open and issue to add real numbers from running on an x86 machine in a future PR? |
@gilbo if the changes look good, we should merge |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Some very slight tweaks. Main thing is we want to pull out those performance numbers before merging the PR
@gilbo addressed comments! |
I was reading the scheduling example and found it a bit magical so trying to rewrite it a bit and add more explanations for how exactly the scheduling operators work and what the goals are. Not ready for review yet so I'll ping when ready.