Ultrasound, X-ray, CT, and PET applications all require intensive back-end compute operations for algorithms such as fast Fourier transform (FFT) using custom pipeline parallelism. Increased algorithm performance is achieved by off loading from the host processor onto an FPGA.
Custom processors created using the OpenCL™1 toolflow are more efficient than multicore CPUs or GPUs, both in computational capability and power requirements.
Informações de produto e desempenho
OpenCL e o logotipo da OpenCL são marcas comerciais da Apple Inc. usadas com permissão pela Khronos.