[FEATURE SUPPORT] add triton gemm kernel

[LoserCheems]

Summary

• Introduces a Triton-kernelized GEMM implementation that enhances matrix updates with alpha/beta scaling, impacting compute-bound workloads.

Root Cause

• The previous implementation lacked an efficient GEMM operation optimized for Triton.

Changes

• Added a new Triton GEMM kernel with autotuning and boundary masking. Updated the README to reflect the new implementation status.

Reproduction

• Not applicable as this is a new feature addition.

Tests

• Existing tests for GEMM were updated to include the new Triton implementation.

Compatibility

• No migration concerns or backwards compatibility issues identified.

Checklist

• Linked issue provided [FEATURE REQUEST] gemm Triton kernel implementation #32
• Adds or updates tests
• Updates docs if needed
• No perf regressions