accel-sim / accel-sim-framework

This is the top-level repository for the Accel-Sim framework.
https://accel-sim.github.io
Other
306 stars 117 forks source link

Methods in adding additional level of cache? #343

Open Tera-NIK opened 1 month ago

Tera-NIK commented 1 month ago

Hi, I am trying to mimic the Composable On-Package GPU (COPA-GPU) paper published by NVIDIA.

The paper adds L3 cache between the HBM and GPU as a form of chiplet.

Any methods on how to implement this structure with Accel-sim, using V100 as basic architecture?