1 / 4

Constant Memory

Constant memory is a specialized memory type in NVIDIA hardware, providing 64 KB of read-only storage accessible by the processor. It is stored in DRAM but can be cached, making it ideal for frequently read data. You can define global variables using `__constant__` and copy data into constant memory from the main memory using `cudaMemcpyToSymbol()`. This guide illustrates how to use constant memory effectively in your kernels with practical examples in CUDA.

chessa
Télécharger la présentation

Constant Memory

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Constant Memory

  2. Constant Memory • NVIDIA hardware provides 64 Kb of constant memory • Read-only by processor • Stored in DRAM but can be cached • Can be useful if something is read repeatedly by the processor

  3. Constant Memory • Define “global” variable with __constant__: __constant__ c_array[N]; • Copy data into the constant memory from main cudaMemcpyToSymbol(c_array, source, size); • Use variable in your kernel

  4. Class Example • Example in class with GA

More Related