Gordon Brown

Computer system architecture trends are constantly evolving to provide higher performance and computing power, to support an increasing demand for high-performance computing domains including AI, machine learning, image processing, and automotive driving aids. One of the most significant being the move towards heterogeneity, where a system has one or more co-processors, often a GPU, working with it in parallel. These kinds of systems are everywhere, from desktop machines and high-performance computing supercomputers to mobile and embedded devices.

This talk will focus on programming GPU co-processors using SYCL; a single-source programming model which allows applications to target GPUs and other accelerators in standard C++. It will first present the GPU architecture and the GPU execution and memory model. It will then present the SYCL programming model and how to use the features of this to efficiently tailor your applications for the GPU including efficient control-flow and how to structure and move your data to achieve efficient utilisation of GPU architectures.