Zero-copy I/O processing for low-latency GPU computing

Cyber-physical systems (CPS) aim to monitor and control complex real-world phenomena where the computational cost and real-time constraints could be a major challenge. Manycore hardware accelerators such as graphics processing units (GPUs) promise to enhancing computation, leveraging the data parallelism often found in real-world scenarios of CPS, but performance is limited by the overhead of the data transfer between the host and the device memory. For example, plasma control in the HBT-EP Tokamak device at Columbia University [11, 18] must execute the control algorithm in a few microseconds, but may take tens of microseconds to copy the data set between the host and the device memory. This paper presents a zero-copy I/O processing scheme that maps the I/O address space of the system to the virtual address space of the compute device, allowing sensors and actuators to transfer data to and from the compute device directly. Experiments using the plasma control system show a 33% reduction in computational cost, and microbenchmarks with more generic matrix operations show a 34% reduction, while in both cases, effective data throughput remains at least as good as the current best performers.