Lots of Data (Vectors)
The Intel® Xeon Phi™ coprocessor is designed with strong support for vector level parallelism with features such as 512-bit vector math operations, hardware prefetching, software prefetch instructions, caches, and high aggregate memory bandwidth. This rich support in hardware coupled with numerous software options may seem overwhelming at first glance, but the richness available offers many solutions for diverse needs. Processors and coprocessors benefit from vectorization. Exposing opportunities to use vector instructions for vector parallel portions of algorithms is the first programming effort. Even then, the layout of data, alignment of data, and the effectiveness of prefetching into caches can affect vectorization. ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access