High-Performance Computing (HPC) refers to hardware-based parallel computation acceleration that delivers performance far beyond the capabilities of standard computer software. Two of the most common technologies used in HPC are Graphics Processing Units (GPUs) and Field Programmable Gate Arrays (FPGAs). GPUs were originally designed for image processing and feature a large number of cores optimized for parallel computation. FPGAs, by contrast, are programmable logic devices that often include dedicated DSP units for hardware-level computation. While GPUs generally provide greater raw computational power, FPGAs offer the unique advantage of integrating customized digital circuits on a single chip, the capabilities not available with GPUs. This makes FPGAs particularly well suited for HPC applications in measurement and control devices, and enables device-on-chip solutions.
Our lab has extended FPGA expertise into HPC applications for medical devices. We are equipped with National Instruments’ LabVIEW FPGA development tools and FlexRIO hardware, providing a robust platform for HPC-driven research.
One of our projects applies FPGA-based HPC to real-time fiber optic monitoring of Spinal Cord Ischemia. The optical system generates a massive volume of data that requires high computational power for time-critical analysis, far beyond what conventional software can deliver. To address this, we implemented the diffusive correlation spectroscopy (DCS) model on FPGA hardware using the LabVIEW FPGA platform. The system is deployed on a PXI chassis, which seamlessly integrates FPGA modules with a PC in a portable configuration. This portable, real-time spinal cord ischemia monitoring system demonstrates the power of FPGA-based HPC and holds promise for a wide range of future medical applications.