December 17, 2020
Los Alamos National Laboratory (LANL), which operates under the purview of the National Nuclear Security Administration (NNSA), is home to a variety of supercom Read more…
April 23, 2015
This week more than 200 New Mexico students and their teachers gathered together at Los Alamos National Laboratory for the 25th annual New Mexico Supercomputing Read more…
April 23, 2014
The 24th annual New Mexico Supercomputing Challenge took place this week at the Los Alamos National Laboratory in Los Alamos, NM. Open to any New Mexico high- Read more…
January 11, 2013
Once intended to save New Mexico from economic uncertainty, now the Encanto supercomputer needs saving. This former number-three superstar is headed to the chopping block. The state is planning to sell off parts of the system to local research universities to recoup some of its investment and pay off outstanding debts. Read more…
The increasing complexity of electric vehicles result in large and complex computational models for simulations that demand enormous compute resources. On-premises high-performance computing (HPC) clusters and computer-aided engineering (CAE) tools are commonly used but some limitations occur when the models are too big or when multiple iterations need to be done in a very short term, leading to a lack of available compute resources. In this hybrid approach, cloud computing offers a flexible and cost-effective alternative, allowing engineers to utilize the latest hardware and software on-demand. Ansys Gateway powered by AWS, a cloud-based simulation software platform, drives efficiencies in automotive engineering simulations. Complete Ansys simulation and CAE/CAD developments can be managed in the cloud with access to AWS’s latest hardware instances, providing significant runtime acceleration.
Two recent studies show how Ansys Gateway powered by AWS can balance run times and costs, making it a compelling solution for automotive development.
Five Recommendations to Optimize Data Pipelines
When building AI systems at scale, managing the flow of data can make or break a business. The various stages of the AI data pipeline pose unique challenges that can disrupt or misdirect the flow of data, ultimately impacting the effectiveness of AI storage and systems.
With so many applications and diverse requirements for data types, management systems, workloads, and compliance regulations, these challenges are only amplified. Without a clear, continuous flow of data throughout the AI data lifecycle, AI models can perform poorly or even dangerously.
To ensure your AI systems are optimized, follow these five essential steps to eliminate bottlenecks and maximize efficiency.
© 2023 HPCwire. All Rights Reserved. A Tabor Communications Publication
Reproduction in whole or in part in any form or medium without express written permission of Tabor Communications, Inc. is prohibited.