Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:


Fujitsu to Showcase Post-K Supercomputer CPU Prototype at ISC 2018

Today Fujitsu unveiled details about their ARM-powered Post K exascale supercomputer, which is planned for deployment at RIKEN in Japan sometime in 2021. Fujitsu has now completed the prototype CPU chip for the machine, which may be the fastest ever commencing functionality field trials. “The Post K processor will likely become the most powerful general-purpose CPU ever for HPC and AI,” said Satoshi Matsuoka from RIKEN on his blog this week.

Welcome to ISC High Performance 2018

In this special guest post, Martin Meuer and Thomas Meuer welcome attendees to ISC 2018, which takes place June 24-28 in Frankfurt, Germany. “Today ISC High Performance stands proudly as a globally recognized HPC event, driven by the extensive involvement of the high performance computing community. The contribution of new topics, speakers, stringent review processes, as well as diversity efforts, makes ISC distinctive.”

Asetek Provides Liquid Cooling QCT at NCHC in Taiwan

Today Asetek announced the deployment of the company’s latest heat rejection technology at the National Center for High-Performance Computing (NCHC) in Hsinchu, Taiwan. “Announced in April, the new liquid cooled NCHC cluster to be implemented by QCT will focus on artificial intelligence. A major goal of the project is to achieve a significant PUE figure, requiring the use of liquid cooling to create a mix of liquid cooled NVIDIA V100 GPUs and Intel Xeon Scalable Processors in a high-density configuration. As one of Asetek’s newest OEM Partners, the NCHC installation by QCT represents both the first major HPC site resulting from the collaboration, and one of the first deployments of the latest InRackCDU.”

Introducing HPE Performance Cluster Manager Software

Over at the HPE Blog, Bill Mannel writes that new HPE Performance Cluster Manager can help you reduce the time and resources spent administering HPC systems. “Managing HPC clusters can be challenging. Especially when you have to deal with the ever-growing large scale systems with thousands of nodes which means you need to monitor and manage more and more components. Your system administrators need a system management solution to allow them to manage even very complex HPC installations easily and quickly.”

Atos and Dell EMC Centre of Excellence coming to Supercomputing Wales

Today Atos announced a new contract with Supercomputing Wales, the national supercomputing research facility for Wales that will result in a world-first Atos and Dell EMC supercomputing Centre of Excellence. This partnership will provide high performance computing facilities and big data research capabilities to a consortium of universities in Wales including Cardiff, Swansea, Aberystwyth and Bangor Universities. “We are thrilled to be part of such an important partnership to help continue the great work already being carried out by universities across Wales,” said Richard Rawcliffe, Vice President, UK Public Sector, Dell EMC.

Prof Michael Resch to present GCS Smart Scale Strategy at ISC 2018

GCS will host a special session on its Smart Scale strategy at ISC 2018. “Prof. Michael Resch will provide insight into the current status and development of Germany’s Smart Scale initiative and of Germany’s position in the field of HPC from an international prospective. The talk takes place Tuesday, June 26 at 12:30pm in room Panorama 2 at the Frankfurt Messe.”

Asetek Liquid Cooling comes to Latest Intel Compute Modules at ISC18

Today Asetek announced plans to showcase its latest liquid cooling technology integrated in Intel Compute Modules at ISC 2018 in Frankfurt. “Asetek is excited with this next step in the ongoing collaboration with Intel,” said John Hamill, Asetek Chief Operating Officer. “The integration of Asetek liquid cooling technology in Intel Compute Modules HNS2600BPBLC and HNS7200APRL serves as further validation of low pressure distributed liquid cooling for HPC and Data Center sites.”

Supermicro, Red Hat, and Solarflare Set World Record STAC Performance

Today Supermicro announced a new world record for lowest latency on the well-known STAC-N1 benchmark. “This benchmark was performed on a pair of Supermicro SYS-1029UX-LL1-S16 servers, each with dual 8-core Intel Xeon Scalable 6144 (Gold) processors overclocked at 4.18GHz. The servers were also loaded with Red Hat’s Red Hat Enterprise Linux 7.5 operating system and Solarflare X2522 Adapters. Compared to all prior publicly released STAC-N1 results, the bare metal system demonstrated the lowest mean latency of 2.3 microseconds at both the base rate (100k messages per second) and the highest rate tested (1 million mps).”

New HPE Pointnext Capabilities to Accelerate Memory-Driven Computing

Today HPE announced the launch of an incubation practice with specialized skills for solving big data problems through Memory-Driven Computing. ” Our vision for Memory-Driven Computing is to enable customers to capture, keep and refine every last bit of their data, up to 10,000 times faster than yesterday’s solutions,” said Beena Ammanath, global vice president, Artificial Intelligence, Data and Innovation, HPE. “The introduction of HPE Pointnext capabilities for Memory-Driven Computing will accelerate our ability to bring Memory-Driven Computing technologies to our customers and help them solve some of their most complex problems and more quickly than ever before.”

Technology Predictions for ISC 2018

In this special guest post, Dr. Rosemary Francis looks forward to the ISC 2018 conference, which starts next week in Frankfurt, Germany. “We recently saw the launch of Summit, the first 200 petaflop system – who will unveil the next giant machine? The race for exascale will certainly to continue to be a topic at ISC and we will be asking the vendors how their plans are progressing.”