Getting Started#
New to GLASS? Start here. GLASS is header-only — there is nothing to build to
use it; add the repo root to your include path and #include "glass.cuh".
These pages get you from a clean checkout to your first working kernel,
explain the single-block execution model the rest of the docs assume, and
list the devices and toolkits the library has been checked on.
Installation
Header-only include model, the CUDA toolkit requirement, and the optional
MathDx setup for the glass::nvidia:: backend.
Library Overview
What GLASS is, the single-block execution model, and the call surfaces — with a guide to choosing the right backend.
Supported Platforms
Which devices and toolkits carry a signed receipt, which have been exercised, and which only compile.