Skip to content

Design space exploration of correlation computation on FPGA.

License

Notifications You must be signed in to change notification settings

fpgasystems/amnes

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Correlation computation [AMNES]

Design space exploration of correlation computation implementation on FPGA.

Project branch structure

  • main - it contains stable project versions
  • dev_"shortcut_name" - it contains developers branch; it does not have to be stable; only its stable versions merge with main branch

Pre-requisites

Repository structure

├── baselines
│   └── gpu       (compute PCC using PyTorch)
│   └── cpp       (sufficient statistics are gathered by the CPU from data stored in the memory)
│   └── outer_lib (extracting the sufficient statistics from the Eigen matrix-matrix multiplication and computing the PCC)
│   └── python    (compute PCC using python's pandas or numpy build in functions) 
├── coyote_integration
│   └── hdl       (RTL *.sv to integrate AMNES as a co-processor kernel into coyote)
│   └── hdl_rdma  (RTL *.sv to integrate AMNES as a network attached kernel into coyote)
│   └── host      (host CPU *.cpp code)
│   └── host_rdma (remote and host CPU *.cpp code)
│   └── testbench (RTL *.sv to test the RTL compiled from Vitis HLS)
├── hls
│   └── class_method (high level synthesis code for Pearson Correlation Coefficient)
├── bitstreams (progammable FPGA files)
    └── co-processor (PCIe data source)
    └── rdma         (RDMA network stack data source)

About

Design space exploration of correlation computation on FPGA.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published