Accelerate Epistasis Detection on Intel® CPUs and Discrete GPUs

This tutorial introduces the basic principles of using Cache-Aware Roofline Model (CARM) to model the performance upper bounds of Intel® CPU and GPU devices. The tutorial uses epistasis detection as a case study, which is an important application in bioinformatics. Using Data Parallel C++ (DPC++) to deploy the application with the Intel® Iris® Xe MAX GPU, it shows how the CARM and Intel® Advisor can be used to detect execution bottlenecks. It also provides useful hints on which type of optimizations to apply to fully exploit both CPU and GPU device capabilities and for use with a hybrid CPU and GPU code design strategy.

Aleksandar Ilic is an assistant professor at the Instituto Superior Técnico, Universidade de Lisboa and a senior researcher of the Instituto de Engenharia de Sistemas e Computadores (INESC-ID). He has contributed to more than 50 scientific publications. His research interests include high-performance and energy-efficient computing and modeling of heterogeneous systems.

Diogo Marques is a member of the High-Performance Computing Architectures and Systems (HPCAS) group at INESC-ID . His research interests include the modeling of multicore and heterogeneous systems. His work helped to improve the accuracy of CARM by proposing the memory metrics and scaled roofs presented in Intel Advisor.

Rafael Campos is a researcher at INESC-ID and part of the HPCAS group. His main interests are performance modeling of heterogeneous systems, with a focus on performance optimization of bioinformatics applications and roofline modeling of high-performance heterogeneous CPU and GPU systems.

Zakhar A. Matveev, PhD, is a principal engineer and a product architect for the Intel Advisor tool on the x86 CPU and Intel GPUs. His focus and professional interests are in the areas of high-computing, parallel programming, hardware and software co-design, and computer graphics.