Skip to content

HeteroCL: A Multi-Paradigm Programming Infrastructure for Software-Defined Heterogeneous Computing

License

Notifications You must be signed in to change notification settings

sqPoseidon/heterocl

 
 

Repository files navigation

GitHub license CircleCI

HeteroCL: A Multi-Paradigm Programming Infrastructure for Software-Defined Reconfigurable Computing

Website | Installation | Tutorials | Samples | Documentation

Introduction

With the pursuit of improving compute performance under strict power constraints, there is an increasing need for deploying applications to heterogeneous hardware architectures with accelerators, such as GPUs and FPGAs. However, although these heterogeneous computing platforms are becoming widely available, they are very difficult to program especially with FPGAs. As a result, the use of such platforms has been limited to a small subset of programmers with specialized hardware knowledge.

To tackle this challenge, we introduce HeteroCL, a programming infrastructure comprised of a Python-based domain-specific language (DSL) and a compilation flow. The HeteroCL DSL provides a clean programming abstraction that decouples algorithm specification from three important types of hardware customization in compute, data types, and memory architectures. HeteroCL can further capture the interdependence among these different customization techniques, allowing programmers to explore various performance/area/accuracy trade-offs in a systematic and productive manner. In addition, our framework currently provides two advanced domain-specific optimizations with stencil analysis and systolic array generation, which produce highly efficient microarchitectures for accelerating popular workloads from image processing and deep learning domains.

Language Overview

flow

Current Compilation Flow

flow

Evaluation on AWS F1 (Xilinx Virtex UltraScale+TM VU9P FPGA)

The speedup is over a single-core single-thread CPU execution on AWS F1.

Benchmark & Data Sizes & Data Type #LUTs #FFs #BRAMs #DSPs Freq. (MHz) CPU Runtime (ms) FPGA Runtime (ms) Speedup
KNN Digit Recognition
K=3 #images=1800
uint49
4.1k (0.42%) 5.5k (0.26%) 38 (2.0%) 0 (0.0%) 250 0.73 0.07 10.4
K-Means
K=16 #elem=320 x 32
int32
168.2k (16.6%) 212.1k (10.0%) 54 (2.8%) 1.5k (22.5%) 187 65.6 0.79 83.0
Jacobi(Stencil)
480x640
fp32
15.2k (1.5%) 26.5k (1.24%) 30 (1.54%) 99 (1.29%) 250 16.27 2.21 7.36
Gaussian(Stencil)
480x640
fp32
28.9k (2.8%) 49.9k (2.4%) 30 (1.54%) 344 (5.1%) 250 20.31 1.6 13.2
Siedel(Stencil)
480x640
fp32
10.2k (0.99%) 19.1k (0.89%) 23 (1.18%) 56 (0.82%) 250 23.86 1.42 16.8

Related Publications

Related Work

HeteroCL is a Python-based DSL extended from TVM and it extends Halide IR for intermediate representation. HeterCL incoporates the SODA framework, PolySA framework, and Merlin Compiler for FPGA back-end generation.

Contributing to HeteroCL

Coding Style (Python)

We follow official Python coding style and use NumPy docstring style.

Coding Style (C and C++)

We follow Google coding style.

Steps

  1. Use clang-format to format your C-related files. The configuration file is in docs/.clang-format. Following is a sample command to format the file in place. Note that you need to put the configuration file at the same directory you execute the command.

    clang-format -i -style=file <cpp-file>

  2. Use Pull Request. Remember to select the most suitable labels and put it in the title.

  3. Make sure all the tests pass.

About

HeteroCL: A Multi-Paradigm Programming Infrastructure for Software-Defined Heterogeneous Computing

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • C++ 75.3%
  • Python 21.0%
  • Makefile 1.2%
  • Java 0.8%
  • CMake 0.6%
  • C 0.5%
  • Other 0.6%