# Portable SIMD 2021-04-12 Meeting

###### tags: `Portable SIMD` `Minutes`

[Last Meeting] | [This Meeting's Chat] | [Next Meeting]

[Last Meeting]: https://hackmd.io/t3zBrB9bQLGkpbL1twBHlw
[This Meeting's Chat]: https://rust-lang.zulipchat.com/#narrow/stream/257879-project-portable-simd/topic/Meeting.202021-04-12/near/234220376
[Next Meeting]: https://hackmd.io/bk7HHiZeSaC7r1tnwFtZhA

## Agenda

0. PR https://github.com/rust-lang/stdsimd/pull/83
1. PR https://github.com/rust-lang/stdsimd/pull/89
2. What should our approach to Wasm SIMD be?
3. Meet on biweekly basis instead?
4. What about aarch64 codegen?
5. BIKESHED TIME: LanesAtMostWhat?

## Resolution

- on Wasm: try to disable the tests that are causing errors specifically, otherwise shrug and ignore the wasm target entirely
- On meeting scheduling: Yes, biweekly or perhaps reschedule
- On aarch64: This should be treated as blocking stable until we understand why.
- On the limiter trait: We don't want to go below 32 lanes. We **very** much don't want to go below 16 lanes.
- What use is a single lane anyways?