Lenovo and Intel unveil 2nd a part of Leibniz Supercomputing Centre’s HPC buildout

Lenovo and Intel unveil 2nd a part of Leibniz Supercomputing Centre’s HPC buildout

cherezoff – stock.adobe.com

On the 2nd ranked because the 15th most highly effective supercomputer on this planet, Lenovo and Intel’s tech will be extinct to up the performance and compute capabilities of the Leibniz Supercomputing Centre’s HPC ambiance

Caroline Donnelly


Published: 17 Jun 2021 14: 22

Efforts to boost the storage and compute capabilities of the Leibniz Supercomputing Centre’s (LRZ) SuperMUC-NG high-performance computing (HPC) setup are location to birth later this year, with the Lenovo Infrastructure Alternate choices Community and Intel to continue their involvement in the venture.

The SuperMUC-NG is at gift ranked the 15th most highly effective supercomputing entity on this planet, and is location to glean a series of enhancements to bolster its capability to help the LRZ person community so as that it’ll handle synthetic intelligence and complicated data analytics workloads more without complications.

Lenovo and Intel were introduced in to work on the venture help in December 2017, confirming at the time that they’d joined forces with LRZ to boost its order to carry out a “subsequent-generation” supercomputer to be utilized by the Bavarian Academy of Sciences in Munich, Germany.

Pegged for completion in 2018, the SuperMUC-NG was as soon as at the foundation configured to bring 26.7 petaflops of compute capability powered by in the case of 6,500 Lenovo server nodes and Intel’s Xeon Platinum processor abilities.

Since polishing off the principle a part of the venture, the water-cooled SuperMUC-NG has been extinct to enact a unfold of projects, together with historical simulation and modelling workloads, besides automating image and sample recognition in planet observations and processing climate data from satellites.

“Given the successful utilisation of SuperMUC-NG in these projects, the request for top-performance data analytics, machine studying and rapid memory performance has extra elevated,” acknowledged Lenovo in an announcement.

To meet this request, the capabilities of the SuperMUC-NG pc are to be expanded thru a 2nd-part buildout of the supercomputer, which is being funded by the Free Verbalize of Bavaria and the German Federal Ministries of Education and Research.

This work will peek the SuperMUC-NG equipped with Intel’s subsequent-generation Xeon Scalable processors and the chipmaker’s upcoming HPC graphics processing unit (GPU) integrated with Lenovo’s ThinkSystem platform abilities. Here’s anticipated to happen in the 2nd quarter of 2022.

As a results of these changes, it is far predicted that its compute nodes will bring four times the performance per watt than the part one setup is at gift able to.

Totally different part two enhancements will consist of the introduction throughout the final quarter of 2021 of a one petabyte asynchronous object storage procedure into the setup so it’ll accommodate a sooner throughput of tidy data volumes.

The firm has additionally confirmed that your total work done during part two of the venture would possibly well well well get a sustainability slant to it, because the aim is to bolster the supercomputer’s performance in a environmentally friendly manner.

As such, it would continue to invent exercise of Lenovo’s liquid-cooling procedure, in the manufacture of its Neptune exclaim water-cooling abilities, which reportedly gets rid of about 90% of the warmth generated by the compute procedure.

“Delivering resources and companies that empower researchers to slip up their projects is at the heart of all the pieces we lift out at LRZ,” acknowledged Dieter Kranzlmüller, LRZ director and professor.

“With Lenovo’s warm-water cooling abilities, we’re ready to bring these enhancements in a manner that is as sustainable and vitality-ambiance friendly as possible.”

Mutter Continues Below

Learn more on Clustering for top availability and HPC

Learn Extra

Leave a Reply

Your email address will not be published. Required fields are marked *