This repo have some implementations of some basic hardware units in RTL language (verilog for now), which can be used for area/power evaluation and support the hardware design tradeoff.
- mux:
- arithmetic_units:
- systolic_array: xinhao
- tensor_core: jun
- N2M_tensor_core: jun
- mac_tree: hongyi
- vector_alu: hongyi
- bitonic_sorting: chiyue
- prefix_sum: kai
- fan: kai
- mrn:
- merge_sort_tree:
- cross_bar: jun/hongyi
- benes_network: hongyi
- set_op_core: zhenhua
Create the repo.
Add original files of these modules:
- systolic_array: xinhao
- bitonic_sorting: chiyue
- prefix_sum: kai
- fan: kai
- Add more modules:
- tensor_core: jun
- mac_tree: hongyi
- vector_alu: hongyi
- Organize the code style.
- Clearify the file structure, input, output, function defination, parameter of each module in code file and readme document.
- Make the plan of parameterize these modules.
- Add more modules:
- crossbar: jun
Make each module parameterized, both in code and document.