• 2019 Update 4
  • 03/20/2019
  • Public Content

Local Memory Usage

One typical GPU-targeted optimization uses local memory for caching of intermediate results. For CPU, all OpenCL™ memory objects are cached by hardware, so explicit caching by use of local memory just introduces unnecessary (moderate) overhead.

Product and Performance Information


Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.