@HPCpodcast: Evaluating Specialty AI Chips at Argonne’s AI Testbed

Our latest episode of @HPCpodcast delves into the Cambrian explosion of AI specialty chips – of which there so many have been released on the HPC-AI market that it’s hard to discern which chip is right for what workload. Hence the AI Testbed at the Argonne Leadership Computing Facility. Shahin and Doug spoke with Venkat Vishwanath….

Exascale: ALCF Aurora Early Adopter Series to Hold Oct. 26 Intro to Intel Extensions of Scikit-learn for ML Frameworks

On Wednesday, Oct. 26 from 11 am to noon CT, the Argonne Leadership Computing Facility’s Aurora Early Adopter series will hold an online event entitled Intro to Intel Extensions of Scikit-learn to Accelerate Machine Learning Frameworks. The event will be led by Bob Chesebrough, Intel solutions architect. Registration is here. Scikit-learn is a library for […]

ALCF Developer Session June 30: Profiling Deep Learning Applications with NVIDIA Nsight

On Thursday, June 30 from 1-2 pm Centeral Time, the Argonne Leadership Computing Facility will hold a developer session on performance analysis techniques for deep learning applications using the NVIDIA Nsight Systems profiling tool to peek under the covers. The session will cover how to collect performance information for the neural network layers to relate […]

Argonne to Host ‘Future of Computing’ May 18

May 16, 2022 — The U.S. Department of Energy’s (DOE) Argonne National Laboratory will host a panel to discuss new paradigms in computing Wednesday, May 18, 10-11:30 a.m. Central Time. Argonne said the event is expected to draw approximately 500 scientists, researchers, engineers and industry professionals. Panelists will discuss new forms of computing for solving pressing […]

Argonne, University Researchers Team to Make Data FAIR for AI

In a new study, researchers from Argonne National Laboratory, Massachusetts Institute of Technology, University of California San Diego, University of Minnesota, and University of Illinois at Urbana-Champaign have laid out new practices to guide the curation of high energy physics datasets that makes them more FAIR — more findable, accessible, interoperable and reusable. Data is the lifeblood […]

Argonne Selects Altair Workload Manager for Polaris, Aurora

Simulation software company Altair has announced that Argonne National Laboratory will utilize Altair PBS Professional workload manager across the organization’s HPC systems at the Argonne Leadership Computing Facility (ALCF)  – including the Polaris and Aurora supercomputers – to accelerate research in science and engineering. PBS Professional – which replaces the ALCF’s in-house Cobalt workload manager […]

Argonne Picks AMD EPYC CPUs for Polaris

Santa Clara — Aug. 30, 2021 — AMD today announced that the U.S. Department of Energy’s (DOE) Argonne National Laboratory has chosen AMD EPYC processors to power the Polaris supercomputer (see earlier coverage), which will prepare researchers for the forthcoming exascale HPC system at Argonne called Aurora. Polaris will use 2nd Gen EPYC processors and then upgrade to 3rd Gen […]

NVIDIA and HPE to Deliver 2,240-GPU Polaris Supercomputer for Argonne

NVIDIA and Argonne National Laboratory this morning announced Polaris, a GPU-based supercomputer, with 2,2240 NVIDIA A100 Tensor Core GPUs delivering 1.4 exaflops of theoretical AI performance and about 44 petaflops of peak double-precision performance, up to 4X faster performance than Argonne’s current supercomputers. Argonne is calling the new system, called “Polaris,” a “testbed” supercomputer to […]

Porting a Particle-in-Cell Code to Exascale Architectures

By Nils Heinonen on behalf of the Argonne Leadership Computing Facility As part of a series aimed at sharing best practices in preparing applications for Aurora, we highlight researchers’ efforts to optimize codes to run efficiently on graphics processing units. Take advantage of upgrades being made to high-level, non-machine-specific libraries and programming models Developed in […]

Reading the Intel Tea Leaves: Pat Gelsinger’s HPC Paradox

As he takes charge of Intel, CEO Pat Gelsinger faces a paradox: his new company is both troubled and a revenue geyser; if Intel is to continue its historical growth rates, he’ll need the skills of a corporate turnaround artist. These contradictions surely apply to Intel’s position in HPC/AI/data center server processors, where the company […]