The project involves research and development of the HPCToolkit performance tools to support profiling and tracing of applications on compute nodes that employ Intel GPUs for accelerators.

The project will consider develop code for OpenCL and Level 0 backends of the Intel OneAPI dpc++ programming model.