Add accumulation triton kernel

[tharapalanivel]

Is your feature request related to a problem? Please describe.

fms-mo's triton matmul kernels already included the emulation of accumulation bits. But lacking of documentations nor examples.

Describe the solution you'd like

A clear and concise example to demonstrate how to use these kernels.

Describe alternatives you've considered

NA

Additional context

Documents should include GPU accumulation bit demo as well.