Moreover, the ever-growing need for performance enhancement leads FPGA-based architectures to scale to multiple devices to benefit from a distributed acceleration. In this context, FPGAs are ideal candidates for such workloads, as they allow design architectures tailored to the stencil regular computational pattern. The computationally intensive essence of ISLs, CNNs, and in general stencil-based workloads, requires solutions able to produce efficient implementations in terms of throughput and power efficiency. Indeed, ISLs apply the same stencil several times until convergence, while CNN layers leverage stencils to extract features from an image. Among the various incarnations of stencil-based computations, Iterative Stencil Loops (ISLs) and Convolutional Neural Networks (CNNs) represent two well-known examples of kernels belonging to the stencil class. Stencil-based algorithms are a relevant class of computational kernels in high-performance systems, as they appear in a plethora of fields, from image processing to seismic simulations, from numerical methods to physical modeling.

Altium designer 19 baidu pan