NEW YORK, October 19, 2022 – Today the ACM, the Association for Computing Machinery and the IEEE Computer Association announced that Marcin Kubik of ETH Zurich and Masado Alexander Ishi of the University of Utah are the recipients of the 2022 ACM-IEEE CS George Michael Memorial HPC Fellowships.
Shelby Lockhart of the University of Illinois at Urbana-Champaign received an honorable mention. Copik is known for integrating the Job-as-a-Service programming model into HPC applications and delivering high performance to serverless servers to lower costs and increase computing efficiency. Ishii has been recognized for developing lightweight, dimensionally-defined parallel entanglement algorithms with a focus on scalability and improving overall time to resolution for engineering applications. Lockhart is recognized for contributions to scalable iterative solutions using node-aware connectivity and low synchronization algorithms to reduce connection bottlenecks on supercomputers.
Kubik’s research bridges the gap between high-performance programming and serverless computing. It brings the Function-as-a-Service (FaaS) programming paradigm to the HPC domain by developing high-performance hardware and software solutions for the serverless stack. By solving key performance challenges for FaaS, it builds a fast and efficient programming model that brings innovative cloud technologies to HPC data centers, allowing users to take advantage of pay-as-you-go billing and helping operators reduce their operating costs and costs. environmental impact.
To this end, he has been working on solutions designed for different levels of the FaaS computing stack, from computers and networks to high-level optimizations and efficient system designs. He has also proposed a new design for serverless platforms that implement HPC practices such as networking with low latency, data location, and efficient communication.
Masado Alexandre Ishii
Ishii is the lead developer of the University of Utah’s Dendro-KT framework for 4D adaptation and parallelism in temporal formulas. Due to the ever-increasing levels of parallelism in the largest machines, parallelism through space is not sufficient – and in many cases, the inability to parallel in time is the biggest bottleneck for many important problems. The Dendro-KT framework addresses this issue and also simplifies the development of time- and space-varying increments, which are important for limiting the computational work required to obtain a given precision.
Working with collaborators, Ishii has also been involved in developing methods and codes for large-scale fluid simulations around complex objects, including a case with multiple complex objects, to assess the risks of COVID-19 transmission in the classroom.
Lockhart made contributions to parallel communications, basic parallel numerical algorithms, and advanced capabilities for large-scale predictive simulation. Her focus has been on modeling performance in heterogeneous settings, with a focus on re-engineering message communications “under the hood” (the aspects of a high-performance architecture that are not easily seen) as well as looking at underlying algorithmic changes in order to dramatically improve achievable performance.
Among the highlights of her research, she provided detailed contact models to drive the choice of message routing, which led to impressive improvements across a range of problem types. It also provided a strategy for achieving impressive reductions in connection costs in GPU systems by communicating through the host, taking into account different data volumes and number of GPUs. In addition, Lockhart’s work on fixed point solutions has made important contributions to the SUNDIALS Project.
About the ACM IEEE CS George Michael Memorial . Fellowship
ACM-IEEE CS George Michael Memorial HPC Fellowship He was honored in memory of George Michael, one of the founders of the SC Conference series. The fellowship honors exceptional doctoral students worldwide whose research focus is on applications of high-performance computing, networking, storage, or large-scale data analytics using the most powerful computers currently available. The fellowship includes a $5,000 award and travel expenses to attend Supreme Committee Conferencewhere fellowships are officially offered.
ACM, Association for Computing Machinery, is the world’s largest educational and scientific computing community, uniting computing educators, researchers, and professionals to inspire dialogue, share resources, and address challenges in the field. ACM strengthens the collective voice of the computing profession through strong leadership, promotion of the highest standards, and recognition of technical excellence. ACM supports the professional growth of its members by providing opportunities for lifelong learning, career development, and professional networking.
SC, International Conference on High Performance ComputingSponsored by ACM and IEEE-CS, it offers a complete technology education program and exhibit to showcase the many ways in which high-performance computing, networking, storage, and analysis are leading to advances in scientific discovery, research, education, and commerce. This first international conference includes a globally attended technical programme, workshops and tutorials, a world-class exhibition area, demonstrations and hands-on learning opportunities.