Add SME support for Apple M4 family of chips

[Developer-Ecosystem-Engineering]

The Apple M4 family of chips include support for SME (Scalable Matrix Extension), which can provide significant performance gains for matrix multiplication. This PR adds support for SME for several types.

Changes:

• Update pytorch/cpuinfo to latest for SME runtime detection
• CMake compiler support detected via sample program using SME instructions
• Added SME support for f32_f32_f32_f32, i8_i8_i32_i32, i8_i8_i32_{i,u}8, u8_u8_i32_u{8,16}

Benchmark results using MacBook Pro with M4 Pro comparing SME and NeonDotProd:

• ruy_benchmark_f32_f32_f32_f32 - up to 14x better
• ruy_benchmark_i8_i8_i32_i32 - up to 7.5x better
• ruy_benchmark_i8_i8_i32_{i,u}8 - up to 7x better
• ruy_benchmark_u8_u8_i32_u{8,16} - up to 7x better
• Other types / sizes have not been optimized and perform similar to NeonDotProd

All tests pass:

• MacBook Pro with M4 Pro (arm64, SME supported)
• MacBook Pro with M3 Max (arm64, no SME support)
• MacBook Pro with M4 (Rosetta x86_64, no SME support)
• iMac Pro (x86_64, no SME support)

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.