The Project Verona allocator, snmalloc, is now available for @rustlang. Great Job https://github.com/SchrodingerZhu/snmalloc-rs … for packaging it up, and @snfernandez for testing it. https://crates.io/crates/snmalloc-rs/ …
-
-
Has anyone looked at comparing mimalloc for producer/consumer allocation workloads? Very interesting... Looking forward to give it a whirl...
@heinz_gies2 replies 0 retweets 2 likes -
Replying to @darachennis @rustlang and
There has not been a recent comprehensive comparison. We borrowed a lot of ideas from mimalloc when it was released to improve our performance: https://github.com/microsoft/snmalloc/issues/67 … Would love to see an independent comparison.
2 replies 1 retweet 2 likes -
Replying to @ParkyMatthew @rustlang and
We have mimalloc in production. Good boost in perf over jemalloc, so that's great given the highlighted changes and our use case - we're happy to feed back anything we learn.
1 reply 1 retweet 4 likes -
-
Replying to @heinz_gies @darachennis and
So here we go, some numbers running mimalloc to our main benchmark, then through the json-bench suite and finally the http://simd-json.rs perf bench:https://gist.github.com/Licenser/24191036f18d95423c12ee96a9fd4ea4 …
3 replies 1 retweet 5 likes -
Replying to @heinz_gies @darachennis and
Updated this with different flag combinations, 1mib seems to be the winner in this use case! :D
3 replies 0 retweets 1 like
Adding @schemeprincess as snmalloc might be a nice perf boost for her crews use case too
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.