@article {, title = {With Extreme Computing, the Rules Have Changed}, journal = {Computing in Science \& Engineering}, volume = {19}, year = {2017}, month = {2017-05}, pages = {52-62}, abstract = {On the eve of exascale computing, traditional wisdom no longer applies. High-performance computing is gone as we know it. This article discusses a range of new algorithmic techniques emerging in the context of exascale computing, many of which defy the common wisdom of high-performance computing and are considered unorthodox, but could turn out to be a necessity in near future.}, doi = {https://doi.org/10.1109/MCSE.2017.48}, author = {Jack Dongarra and Stanimire Tomov and Piotr Luszczek and Jakub Kurzak and Mark Gates and Ichitaro Yamazaki and Hartwig Anzt and Azzam Haidar and Ahmad Abdelfattah} } @inproceedings {959, title = {Weighted Dynamic Scheduling with Many Parallelism Grains for Offloading of Numerical Workloads to Multiple Varied Accelerators}, journal = {Proceedings of the 6th Workshop on Latest Advances in Scalable Algorithms for Large-Scale Systems (ScalA{\textquoteright}15)}, volume = {No. 5}, year = {2015}, month = {2015-11}, publisher = {ACM}, address = {Austin, TX}, abstract = {A wide variety of heterogeneous compute resources are available to modern computers, including multiple sockets containing multicore CPUs, one-or-more GPUs of varying power, and coprocessors such as the Intel Xeon Phi. The challenge faced by domain scientists is how to efficiently and productively use these varied resources. For example, in order to use GPUs effectively, the workload must have a greater degree of parallelism than a workload designed for a multicore-CPU. The domain scientist would have to design and schedule an application in multiple degrees of parallelism and task grain sizes in order to obtain efficient performance from the resources. We propose a productive programming model starting from serial code, which achieves parallelism and scalability by using a task-superscalar runtime environment to adapt the computation to the available resources. The adaptation is done at multiple points, including multi-level data partitioning, adaptive task grain sizes, and dynamic task scheduling. The effectiveness of this approach for utilizing multi-way heterogeneous hardware resources is demonstrated by implementing dense linear algebra applications.}, keywords = {dataflow scheduling, hardware accelerators, multi-grain parallelism}, author = {Azzam Haidar and Yulu Jia and Piotr Luszczek and Stanimire Tomov and Asim YarKhan and Jack Dongarra} }