Bambu: an Open-Source Research Framework for the High-Level Synthesis of Complex Applications
Event Type
Special Session (Research Track)
Virtual Programs
Hosted in Virtual Platform
Machine Learning/AI
DescriptionThis talk presents the open-source high-level synthesis research framework Bambu.
The framework provides an open-source starting point to experiment with new ideas that span high-level synthesis, high-level verification and debugging, FPGA/ASIC design, platform integration, design space exploration, and parallelized hardware accelerator design.
The tool accepts as input standard C/C++ specifications as well as compiler intermediate representations (IRs) coming from well-known Clang/LLVM and GCC compilers.
This allows the EDA research community to integrate new transformations on the input sources as well as on the compiler IRs.
The easy extendable modular framework comes with many already implemented optimizations together with many HLS benchmarks.
The integration with synthesis/verification commercial/open-source backends allows the researchers to easily test any new finding and easily get metrics with respect to the performance and the resource usage required by a given application. Different FPGA devices are already supported: AMD/XILINX, Intel/Altera, Lattice Semiconductor, and NanoXplore FPGAs. Finally, the integration with the OpenRoad open-source end-to-end silicon compiler perfectly fits very well with the recent push towards open-source EDA.