4 ms·

The issue with this is that if you care about speed, you need to be using SIMD -- and wasm doesn't seem to want to support it -- stating a "minimum viable" stan

by ezy 11y ago

The issue with this is that if you care about speed, you need to be using SIMD -- and wasm doesn't seem to want to support it -- stating a "minimum viable" standard from the year 2000.

jacobolus 11y ago
Existing Javascript JITs also don’t output SIMD instructions as far as I know. I expect SIMD support can be added to wasm at some future date.
- ezy 11y ago
  I wasn't really referring to wasm vs JS (which is easy to outperform), but reaching equivalent speed to native, which is the ultimate goal, I'd hope. Otherwise you're just throwing performance away... But, I was a little quick on the trigger. It's not obvious, but after drilling through the design docs for a half hour.. I found a tentative discussion of how SIMD might be added later (based on some kind extension mechanism).
nwmcsween 11y ago
Why wouldn't llvm be able to emit SIMD when it can vectorize? The idea the SIMD is the end all to fast computing is pretty short sighted IMO.
- jacobolus 11y ago
  SIMD isn’t the “end all”, but upcoming Intel chips have instructions for handling 8 doubles or 16 floats per instruction, and if you’re trying to implement a video codec or large-scale physics simulation, an order of magnitude of speed difference can make or break your app. To take an example where timeliness is crucial, think of the difference between, say, 10 frames per second vs. 60 frames per second in a first-person shooter game.
azakai 11y ago
SIMD.js is in progress for JS, so JS VMs are already working on it. It is on the roadmap for being added to WebAssembly, likely with a similar API.
- zurn 11y ago
  Is there a technical restriction why JS JITs don't transparently apply SIMD to existing loops like normal compilers (=autovectorization)? Or is it in the cards to build it on top of SIMD.js?
  - yoklov 11y ago
    Autovectorization is hard and fragile in the best of times -- tight stable loops with no (or few) branches and little/no memory access that may be aliased, etc. For a dynamic language JIT this is pretty much infeasible (as I understand it). Every loop might have branches for guards/bailing out due to deopts, and at least in JS, TypedArrays are allowed to alias eachother.

by ezy 11y ago

The issue with this is that if you care about speed, you need to be using SIMD -- and wasm doesn't seem to want to support it -- stating a "minimum viable" standard from the year 2000.

jacobolus 11y ago

Existing Javascript JITs also don’t output SIMD instructions as far as I know. I expect SIMD support can be added to wasm at some future date.

ezy 11y ago
I wasn't really referring to wasm vs JS (which is easy to outperform), but reaching equivalent speed to native, which is the ultimate goal, I'd hope. Otherwise you're just throwing performance away... But, I was a little quick on the trigger. It's not obvious, but after drilling through the design docs for a half hour.. I found a tentative discussion of how SIMD might be added later (based on some kind extension mechanism).

nwmcsween 11y ago

Why wouldn't llvm be able to emit SIMD when it can vectorize? The idea the SIMD is the end all to fast computing is pretty short sighted IMO.

jacobolus 11y ago
SIMD isn’t the “end all”, but upcoming Intel chips have instructions for handling 8 doubles or 16 floats per instruction, and if you’re trying to implement a video codec or large-scale physics simulation, an order of magnitude of speed difference can make or break your app. To take an example where timeliness is crucial, think of the difference between, say, 10 frames per second vs. 60 frames per second in a first-person shooter game.

azakai 11y ago

SIMD.js is in progress for JS, so JS VMs are already working on it. It is on the roadmap for being added to WebAssembly, likely with a similar API.

zurn 11y ago
Is there a technical restriction why JS JITs don't transparently apply SIMD to existing loops like normal compilers (=autovectorization)? Or is it in the cards to build it on top of SIMD.js?
- yoklov 11y ago
  Autovectorization is hard and fragile in the best of times -- tight stable loops with no (or few) branches and little/no memory access that may be aliased, etc. For a dynamic language JIT this is pretty much infeasible (as I understand it). Every loop might have branches for guards/bailing out due to deopts, and at least in JS, TypedArrays are allowed to alias eachother.