Should the academic community launch an open-source FPGA device and tools effort?: evening panel.

ABSTRACTFor years, many academic researchers in reconfigurable computing have been frustrated by their reliance on commercial FPGAs and tools. Commercial FPGAs have highly complex micro-architectures, come with undocumented binary interfaces, have no compatibility between generations, and come with difficult to use proprietary place and route tools. The FPGA vendors are making the right moves for serving their commercial customer base, but it seems at times these moves are in conflict with the needs of the academic research community. These problems make it difficult for academics to teach FPGA design and to participate in relevant research related to FPGAs. Complex FPGA fabrics and undocumented proprietary interfaces make it nearly impossible to build new tools to target actual commercial devices. University researchers are now hesitant to participate in FPGA architecture research because the level of complexity of FPGAs is beyond what is possible to implement in an academic setting. Reconfigurable computing application developers are at the mercy of complex FPGA devices and closed-source place and route tools - not optimized for computing applications. Some argue that the time is right for academia to break away and create an open source FPGA device. Such an effort would enable university researchers to take an active role in architecture design, and promote an open source community effort in developing and advancing the state of the art of tools. A similar situation existed in the early 80's surrounding processor design. As a result of competition among the big computer companies, microprocessor designs had evolved from very simple beginnings to highly complex architectures and equally complex compilers. Universities were largely on the sidelines, as they couldn't compete at the level of industrial designs of the day. "RISC" changed all that. By adopting simple architectures, universities were once again able to actually design and implement their own processors and to innovate with microarchitecture and compilers. Is it time for a "RISC" revolution for FPGAs? FPGA companies might argue that FPGA platforms are necessarily complex and out of the realm of what's possible to implement within an academic setting, and would say that an open source effort in not necessary, as they are helping to promote academic research by collaborations with universities, and donation of tools, and hardware platforms. This panel will discuss the pros and cons of an academic community led open FPGA device and tools development effort.