GPUs can be used for much more than graphics processing. As opposed to a CPU, which can only run four or five threads at once, a GPU is made up of hundreds or even thousands of individual, low-powered cores, allowing it to perform thousands of concurrent operations. Because of this, GPUs can tackle large, complex problems on a much shorter time scale than CPUs. Dive into parallel programming on NVIDIA hardware with CUDA Succinctly by Chris Rose, and learn the basics of unlocking your graphics card.
Topics included: Introduction • Creating a CUDA Project • Architecture • First Kernels • Porting from C++ • Shared Memory • Blocking with Shared Memory • NVIDIA Visual Profiler (NVVP) • Nsight • CUDA Libraries.
Download Free PDF / Read Online
Publisher: Syncfusion Inc.
Published: January 2015
Format(s): PDF, Mobi(Kindle)
File size: 3.13 MB
Number of pages: 119
Download / View Link(s): PDF, Mobi