[feat] Int8 Matmul Training kernel

Nobody has claimed this yet.

### 🚀 The feature, motivation and pitch

W8A8 (int8 for both weight and activation) matmul is beneficial for A100 and could provide great memory and speed benefits, and could be effective given dedicated design on mix precision training (avoid convergence issue and mitigate performance drop). The kernel will be a fused int8 matmul kernel with fusion of three parts:

input activation quantization to int8 (dynamic, static) + int8 matmul (weight int8 dynamic, static quantization inside) + output dequantization

### Alternatives

_No response_

### Additional context

_No response_

The issue describes a fused W8A8 int8 matmul training kernel combining activation quantization, int8 matmul, and output dequantization, but names no files, tests, or entry points. Start by locating the existing training-kernel implementation and clarifying dynamic versus static quantization requirements, target A100 behavior, and acceptance tests; done should include the fused kernel with validated convergence and performance.

Written by the indexing model from the issue text.