Enable platform agnostic SIMD support for vector

[JulieLeeMSFT]

Overview

Part of the library side work here is ensuring that users looking to support ARM64 can do so trivially even when they aren't familiar with the platform and to reduce overall cost of needing to support multiple platforms with SIMD acceleration (x86, x64, and ARM64 today; potentially other platforms like WASM in the future).

That work is being tracked by: #49397 which tracks us exposing a common cross-platform API surface that allows users to utilize "write-once" algorithms and successfully support multiple architectures.

Work Items

• Enable platform agnostics SIMD
  • Expose cross-platform helpers for Vector64, Vector128, and Vector256 #49397
  • monoVM support: [Mono] Intrinsify API's of System.Numerics.Vector and new API's of System.Runtime.Intrinsics.Vector{64, 128} on Arm64 #64072
  • JIT support: Additional cross-platform hardware intrinsic APIs for loading/storing, reordering, and extracting a per-element "mask" #63331

Benchmarks to use

• Microbenchmarks

What Success Looks Like

• The new API does not degrade the performance of the platform specific SIMD.
• Advertise the feature to the community, e.g., through blog posts or twitter.

[dotnet-issue-labeler]

I couldn't figure out the best area label to add to this issue. If you have write-permissions please help me learn by adding exactly one area label.

[ghost]

Tagging subscribers to this area: @dotnet/area-meta
See info in area-owners.md if you want to be subscribed.

Issue Details

---

Overview

Part of the library side work here is ensuring that users looking to support ARM64 can do so trivially even when they aren't familiar with the platform and to reduce overall cost of needing to support multiple platforms with SIMD acceleration (x86, x64, and ARM64 today; potentially other platforms like WASM in the future).

That work is being tracked by: #49397 which tracks us exposing a common cross-platform API surface that allows users to utilize "write-once" algorithms and successfully support multiple architectures.

Work Items

• Enable platform agnostics SIMD
  • Expose cross-platform helpers for Vector64, Vector128, and Vector256 #49397
  • monoVM support: [Mono] Intrinsify API's of System.Numerics.Vector and new API's of System.Runtime.Intrinsics.Vector{64, 128} on Arm64 #64072
  • JIT support: Additional cross-platform hardware intrinsic APIs for loading/storing, reordering, and extracting a per-element "mask" #63331

Benchmarks to use

• Microbenchmarks

What Success Looks Like

• The new API does not degrade the performance of the platform specific SIMD.
• Advertise the feature to the community, e.g., through blog posts or twitter.

Author:	JulieLeeMSFT
Assignees:	tannergooding, JulieLeeMSFT
Labels:	area-Meta, User Story, Priority:2, Cost:L
Milestone:	7.0.0

[dakersnar]

Everything planned here for 7.0 has been completed, with the exception of the details in #63331 that were moved out to 8.0.