Performance is critical in solutions Things-enabled factories, as it enables greater analysis of product quality, equipment performance, and other factors. Overall performance is up 15 percent in the new processors, while signal and image processing get an additional 2x boost with Intel Advanced Vector Extensions (Intel AVX) 2.0. As shown in, Intel AVX 2.0 introduces a fully pipelined Fused Multiply Add (FMA3) that provides twice embedded systems performance of the previous generation for the multiply-add workloads common in image processing. Intel AVX 2.0 also extends integer vector processing from 128 bits to 256 bits, also doubling throughput for many fanless embedded systems workloads. Along the way we’ll present examples where members of the Intel Intelligent Systems Alliance are taking advantage of the new embedded systems.
refer to:
沒有留言:
張貼留言