Investigate and Integrate Experimental SIMD Optimizations from Go 1.26 for Performance Improvements

[gitctrlx]

With the introduction of experimental SIMD optimizations in Go 1.26, this issue aims to:

• Evaluate the potential performance improvements for the gomat library by leveraging the new Go 1.26 SIMD support.
• Design and implement benchmarks comparing the current state against the SIMD-optimized code paths.
• Ensure compatibility with existing BLAS/LAPACK acceleration features, including Intel MKL/OpenBLAS.
• Document and highlight cases where SIMD provides significant benefits or limitations.

Tasks:

• Research Go 1.26 SIMD features including guidelines, usage, and limitations
• Prototype SIMD-enabled functions within gomat
• Analyze performance data and report findings
• Plan for a stable integration strategy if experimental results are promising

Relevant links:

• Go 1.26 Release Notes
• Go SIMD Proposal

If you are interested in performance optimization at the library level, please join the discussion and contribute insights or code.