RapidStream TAPA

RapidStream TAPA is a powerful framework for designing high-frequency FPGA dataflow accelerators. It combines a powerful C++ API for expressing task-parallel designs with advanced optimization techniques from RapidStream to deliver exceptional design performance and productivity.

High-Frequency Performance: Achieve 2× higher frequency on average compared to Vivado¹.
Rapid Development: 7× faster compilation and 3× faster software simulation than Vitis HLS².
Expressive API: Rich C++ syntax with dedicated APIs for complex memory access patterns and explicit parallelism.
HBM Optimizations: Automated design space exploration and physical optimizations for HBM FPGAs.

Quick Start

Prerequisites

Ubuntu 18.04+, Debian 10+, RHEL 9+, Fedora 34+, or Amazon Linux 2023
Vitis HLS 2022.1 or later

Installation

# Install dependencies (Ubuntu/Debian example)
sudo apt-get install g++ iverilog ocl-icd-libopencl1

# Install RapidStream TAPA
# You may rerun this command to update to the latest version.
sh -c "$(curl -fsSL tapa.rapidstream.sh)"

# Optional: Install RapidStream
sh -c "$(curl -fsSL rapidstream.sh)"

Compilation

git clone https://github.com/rapidstream-org/rapidstream-tapa
cd rapidstream-tapa/tests/apps/vadd

# Software simulation
tapa g++ vadd.cpp vadd-host.cpp -o vadd
./vadd

# Hardware compilation and emulation
tapa compile \
   --top VecAdd \
   --part-num xcu250-figd2104-2L-e \
   --clock-period 3.33 \
   -f vadd.cpp \
   -o vecadd.xo
./vadd --bitstream=vecadd.xo 1000

Optimize with RapidStream

rapidstream-tapaopt \
  --work-dir ./build \
  --tapa-xo-path vecadd.xo \
  --device-config u55c_device.json \
  --autobridge-config floorplan_config.json

For detailed instructions, see our User Guide.

Documentation

Success Stories

Serpens (DAC'22): 270 MHz on Xilinx Alveo U280 HBM board with 24 HBM channels, while the Vitis HLS baseline failed in routing.
Sextans (FPGA'22): 260 MHz on Xilinx Alveo U250 board with 4 DDR channels, while the Vivado baseline achieves only 189 MHz.
SPLAG (FPGA'22): Up to 4.9× speedup over state-of-the-art FPGA accelerators, up to 2.6× speedup over 32-thread CPU running at 4.4 GHz, and up to 0.9× speedup over an A100 GPU.
AutoSA Systolic-Array Compiler (FPGA'21): Significant frequency improvements over the Vitis HLS baseline.
KNN (FPT'20): 252 MHz on Xilinx Alveo U280 board, compared to 165 MHz with the Vivado baseline.

Licensing

RapidStream TAPA is open-source software licensed under the MIT license. For full license details, please refer to the LICENSE file. By contributing to this project, you agree to the RapidStream Contributor License Agreement. For more information, please read the full CLA.

Publications

Licheng Guo, Yuze Chi, Jie Wang, Jason Lau, Weikang Qiao, Ecenur Ustun, Zhiru Zhang, Jason Cong. AutoBridge: Coupling coarse-grained floorplanning and pipelining for high-frequency HLS design on multi-die FPGAs. FPGA, 2021. (Best Paper Award)
Yuze Chi, Licheng Guo, Jason Lau, Young-kyu Choi, Jie Wang, Jason Cong. Extending high-level synthesis for task-Parallel programs. FCCM, 2021.
Young-kyu Choi, Yuze Chi, Jason Lau, Jason Cong. TARO: Automatic optimization for free-running kernels in FPGA high-level synthesis. TCAD, 2022.
Licheng Guo, Pongstorn Maidee, Yun Zhou, Chris Lavin, Eddie Hung, Wuxi Li, Jason Lau, Weikang Qiao, Yuze Chi, Linghao Song, Yuanlong Xiao, Alireza Kaviani, Zhiru Zhang, Jason Cong. RapidStream 2.0: Automated parallel implementation of latency insensitive FPGA designs through partial reconfiguration. TRETS, 2023.
Licheng Guo, Yuze Chi, Jason Lau, Linghao Song, Xingyu Tian, Moazin Khatti, Weikang Qiao, Jie Wang, Ecenur Ustun, Zhenman Fang, Zhiru Zhang, Jason Cong. TAPA: A scalable task-parallel dataflow programming framework for modern FPGAs with co-optimization of HLS and physical design. TRETS, 2023.