Embedded heterogeneous programming with OpenCL
Abstract
Modern embedded processors often include graphical processing units (GPUs) that have recently gained to the ability to be programmed with general purpose code. This has resulted in the concept of “heterogeneous programming,” where a program must be conceptualized as separate programs running on processors of different technologies, with one processor controlling the other as a peripheral.
As with other processor types, GPUs derive their performance by using a combination of multicore with single instruction, multiple data parallelism. However, because GPUs place more emphasis on data-level parallelism, GPUs have allowed a new programming model to evolve in which data elements ...
Get Embedded Systems now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
