0.5 performance

624 views
Skip to first unread message

David Anthoff

unread,
Apr 22, 2016, 2:14:12 PM4/22/16
to juli...@googlegroups.com

Should non-core devs start to look at 0.5 performance, or is it all still too much a work in progress to report back how performance for existing code bases changes between 0.4 and 0.5?

 

Just for fun I ran one of my code bases today on 0.4 and on 0.5. I see a pretty significant increase in runtime on 0.5, about 35%:

 

julia 0.4: 665.095440 seconds (2.03 G allocations: 36.515 GB, 0.53% gc time)

julia 0.5: 904.061203 seconds (4.55 G allocations: 75.579 GB, 0.72% gc time)

 

Also, is there a preference from the core team on how to report this kind of stuff? I don’t have the time nor expertise to track down the root cause of what I’m seeing there, and the code I’m using is not public at this point. I’d be happy to give julia core devs access to the repo if someone wanted to investigate what is going on.

 

Cheers,

David

 

--

David Anthoff

University of California, Berkeley

 

http://www.david-anthoff.com

 

Stefan Karpinski

unread,
Apr 22, 2016, 2:23:34 PM4/22/16
to juli...@googlegroups.com
Definitely! Reports of performance regressions are incredibly valuable. You can just open an issue titled with "perf regression: $(description)" and some code to reproduce if possible. We're very much in the phase of trying to take care of these.

Viral Shah

unread,
Apr 23, 2016, 10:12:29 AM4/23/16
to julia-dev
+1

This is definitely the right time to start testing codes with 0.5 and report regressions. In my perfect world, we would have 0.5 at JuliaCon!

-viral

David Anthoff

unread,
Apr 25, 2016, 6:38:53 PM4/25/16
to juli...@googlegroups.com

Alright, I got it in shape so that one can run the example. Here is the issue:

 

https://github.com/JuliaLang/julia/issues/16047

 

The code is not public, but I’m happy to give access to any core developer who wants to figure out what is going on.

Reply all
Reply to author
Forward
0 new messages