Design and Evaluation of Performance-efficient SoC-on-FPGA for Cloud-based Healthcare Applications

Cloud computing with more resources at disposal has served heavy computing demands for various applications. Health is one such sector where the cloud services have benefited by not only storing and updating the subscribers' health records frequently but also providing useful analytical and predictive information to the users. However most of the cloud processing runs are established on general-purpose computing devices that are non-customized for accelerating medical inferences. Hence, a special purpose and power-efficient SoC design for cloud computing targeted towards healthcare applications is desirable. The study examines the design and evaluation of SoCs for today's neural networks trained for healthcare data. The work realizes fully connected neural networks comprised of 2 and 3 layers that are trained from two freely available datasets; one meant for detecting the survival status of the individuals suffering from prostate cancer and the other for survival status post blood and marrow transplantation, respectively. Additionally, a light-weight image classifier derived from medical based MNIST pathological dataset was also considered. The three neural networks were hardware designed and characterized with two different processor cores and two extreme cache memory configurations individually and in combination to generate multiple SoC designs. Each SoC design is illustrated and evaluated in terms of power and hardware resource utilization. The SoC design incorporating all three hardware accelerators compared to the existing cloud platforms showcased power reduction by 7.13X, 5.25X, and 15X, and average throughput improvement by 49.37X, 19.78X, and 26.47X, when compared to its corresponding neural network, runs on CPU, GPU, and TPU units respectively. The SoC designs in this work with three different co-processor choices set an example to achieve customized cloud SoC designs that are power and performance efficient and targeted for healthcare applications.