FPGA Architecture for Computing

Despite their promise in both processing performance and efficiency, today’s FPGAs are woefully inadequate as computing devices, particularly due to their lack of a standard native memory architecture. In a major departure from conventional FPGAs (which are designed to compete with ASIC technologies) the CoRAM (Connected RAM) project endeavors to completely rethink the FPGA architecture from scratch to create a new generation of first-class reconfigurable computing devices. The goal of the CoRAM project is to harness the full potential of FPGAs for computing by introducing a proper memory architecture that serves as a portable, high-performance bridge between the distributed in-fabric computation kernels and the external off-chip memory. In support of this architecture, the CoRAM project will also study and develop the associated mechanisms, from the microarchitecture-level down to the circuit-level, using a combination of FPGA-based prototypes and test chips. This is a joint project with Prof. Ken Mai. Funding for this work is provided, in part, by the National Science Foundation (CCF-1012851) and by Altera. We thank Altera and Xilinx for their donation of tools and hardware. Please see the CoRAM project page for complete and up-to-date details.

  • Students
  • Demo and Downloads
  • Publications
    • ShrinkWrap: Compiler-Enabled Optimization and Customization of Soft Memory Interconnects. Eric Chung and Michael Papamichael. Proc. IEEE Symposium on Field-Programmable Custom Computing Machines (FCCM), April 2013.
    • C-To-CoRAM: Compiling Perfect Loop Nests to the Portable CoRAM Abstraction. Gabriel Weisz and James C. Hoe. Proc. ACM International Symposium on Field-Programmable Gate Arrays (FPGA), February 2013. (pdf)
    • Memory Bandwidth Efficient Two-Dimensional Fast Fourier Transform Algorithm and Implementation for Large Problem Sizes. Berkin Akın, Peter A. Milder, Franz Franchetti and James C. Hoe. Proc. IEEE International Symposium on Field-Programmable Custom Computing Machines (FCCM), April 2012. (pdf)
    • CONNECT: Re-Examining Conventional Wisdom for Designing NoCs in the Context of FPGAs. Michael Papamichael and James C. Hoe. Proc. ACM International Symposium on Field-Programmable Gate Arrays (FPGA), February 2012. (pdf)
    • Prototype and Evaluation of the CoRAM Memory Architecture for FPGA-Based Computing. Eric S. Chung, Michael K. Papamichael, Gabriel Weisz, James C. Hoe, and Ken Mai. Proc. ACM International Symposium on Field-Programmable Gate Arrays (FPGA), February 2012. (pdf)
    • CoRAM: An In-Fabric Memory Architecture for FPGA-Based Computing. Eric S. Chung. PhD Thesis, August 2011. (pdf)
    • CoRAM: An In-Fabric Memory Architecture for FPGA-based Computing. Eric S. Chung, James C. Hoe, and Kenneth Mai. Proc. ACM International Symposium on Field-Programmable Gate Arrays (FPGA), pp 97~106, February 2011. (pdf)
    • Single-chip Heterogeneous Computing: Does the future include Custom Logic, FPGAs, and GPUs? Eric S. Chung, Peter A. Milder, James C. Hoe, and Kenneth Mai. Proc. ACM/IEEE International Symposium on Microarchitecture (MICRO), pp 53~64, December 2010. (pdf)
  • Related