Presentations and other assets from this event are presented here. For information on upcoming events, click here.
Parallel computer processors have become ubiquitous. Today, everything
from a mobile phone, to a desktop computer, to a supercomputer, supports
parallel programming. Despite the adoption of parallelism by processor manufacturers,
few software applications today can use modern processors to their full
potential due to the complexity of parallel algorithm design and software
The Fields Institute will host a three day workshop on parallel computing
led by an expert in the field that will provide trainees with a mixture
of theory and hands-on experience. Trainees will learn parallel programming
with OpenCL, and so will be able to apply their knowledge to CPUs, GPUs,
FPGAs, and other accelerators such as the Intel Xeon Phi.
Understand why parallel programming is important and what makes it hard
Understand the hardware trends that drive parallel programming
Learn the role of the OpenCL specification in the processor and software
Understand the theoretical limits of parallel computing
NC = P?
Universality of consensus and the importance of atomic instructions
Learn how to write OpenCL software for heterogeneous computing
Learn the mathematical foundations of the OpenCL 2.0 memory model
Learn how to write parallel software for CPUs, FPGAs, GPUs, and other