Issue #16487 has been updated by byroot (Jean Boussier).


> It should be decided based on not micro benchmark

That's understandable. I'll try to produce a proper macro benchmark against=
 https://github.com/Shopify/ruby/pull/2. I also agree that Rails apps are p=
robably the best thing to benchmark, as they're quite string heavy.

> How to coordinate configure and ifdefs. What is the default behavior? How=
 to specify to use SIMD?

I started working on that a bit: https://github.com/ruby/ruby/commit/2e9198=
a3af6eb8df8c9ebe40c4bb23b7e30a953d

The idea is to have different functions, and to select which implementation=
 to use based on CPUID.

----------------------------------------
Misc #16487: Potential for SIMD usage in ruby-core
https://bugs.ruby-lang.org/issues/16487#change-83849

* Author: byroot (Jean Boussier)
* Status: Open
* Priority: Normal
* Assignee: =

----------------------------------------
### Context

There are several ruby core methods that could be optimized with the use of=
 SIMD instructions.

I experimented a bit on `coderange_scan` https://github.com/Shopify/ruby/pu=
ll/2, and Pavel Rosick=FD experimented on `String#strip` https://github.com=
/ruby/ruby/pull/2815.

### Problem

The downside of SIMD instructions is that they are not universally availabl=
e.
So it means maintaining several versions of the same code, and switching th=
em either statically or dynamically.

And since most Ruby users use precompiled binaries from repositories and su=
ch, it would need to be dynamic if we want most users to benefit from it.

So it's not exactly "free speed", as it means a complexified codebase.

### Question

So the question is to know wether ruby-core is open to patches using SIMD i=
nstructions ? And if so under which conditions.

cc @shyouhei





-- =

https://bugs.ruby-lang.org/

Unsubscribe: <mailto:ruby-core-request / ruby-lang.org?subject=3Dunsubscribe>
<http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>