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


13.7 Petaflop MareNostrum 4 Supercomputer Coming to BSC

Today the Barcelona Supercomputing Center announced plans to MareNostrum 4, a 13.7 Petaflop supercomputer that will be 12.4 times more powerful than the current MareNostrum 3 system. In a contract valued at almost €30 million, IBM will integrate in one sole machine using its own technologies alongside those of Lenovo, Intel, and Fujitsu.

Avere Speeds Hybrid Cloud at SC16

“Avere Hybrid Cloud is a groundbreaking solution that changes the economics and functionality of data storage and computing. Avere Hybrid Cloud enables companies to leverage public and private storage clouds, IaaS, and elastic compute for workloads previously limited to on-premises high-performance (HPC) storage systems. With Avere Systems, you can achieve unlimited performance scaling, eliminate latency and cut storage costs by more than 50%.”

HPE Goes All-Flash for Public Cloud with 3PAR, Lowers Long Term Storage Costs with Spectra Logic

Today HPE announced today a comprehensive initiative that brings together the best of on-premises performance, application availability, and control with the convenience and agility of public cloud consumption models.

SGI and DDN Power UK Met Office SPICE System for Weather and Climate Research

SGI, Bright Computing, and DDN recently announced that the UK Met Office has selected the three HPC vendors to provide HPC for its new Scientific Processing and Intensive Compute Environment (SPICE) system. SPICE will enable weather and climate researchers to dramatically reduce time required to analyze massive amounts of climate simulation data.

RSC Tornado Breaks Record with 1.41 Petaflops per Rack

Today Russia’s RSC Group announced that the company has achieved a record compute density of 1.41 Pflops per rack using direct liquid cooling and Intel Xeon Phi processors. “RSC supported the TUM student team from Munich with 8 nodes mobile cluster based on RSC Tornado direct liquid cooled architecture. This computing system provided stable operation of computing nodes in “hot water” mode at +63 °С cooling agent temperature at node inputs and had the following configuration: 72-cores Intel Xeon Phi 7290 processors, Intel S7200AP server boards, Intel SSD DC S3500 Series M.2 340 GB solid-state drives, switch and adapters based on Intel Omni-Path high-speed fabric, highly efficient Micron DDR4-2400 VLP 16-32 GB memory modules.”

BSC Paper: Scale-in on Large-memory Nodes to Save Energy

Researchers from the Barcelona Supercomputing Center have received the award for Best Paper of the MEMSYS 2016 conference. “This study confirms that improvements in the memory system can have significant impact on the real world, improving power and energy, performance, and/or dollar cost. The MEMSYS program committee liked the study especially because, in contrast to most of the prior research, it can be applied immediately in the production systems, and it requires no changes to the system architecture, Operating System, system software or applications.”

Huawei Showcases FusionServer X6000 at SC16

“Huawei’s new-generation FusionServer X6000 HPC platform, based on the Intel Scalable System Framework, provides the balanced performance, scalability, and flexibility to efficiently handle a wide range of HPC workloads,” said Hugo Saleh, Director of Marketing and Enabling, Intel Scalable Datacenter Solutions. “The FusionServer X6000 is a great example of how Intel Xeon processors and Intel Omni-path architecture can combine to build industry leading solutions with innovative fabric features that optimize performance, resiliency, and traffic movement.”

Liquid Silicon Could Bridge the Gap Between Computation and Storage

Researchers at the University of Wisconsin–Madison are developing new computer chips that combine tasks usually kept separate by design. According to assistant professor Jing Li, these “liquid silicon” chips can be configured to perform complex calculations and store massive amounts of information within the same integrated unit — and communicate efficiently with other chips. “There’s a huge bottleneck when classical computers need to move data between memory and processor,” says Li. “We’re building a unified hardware that can bridge the gap between computation and storage.”

Manage Your Lustre HPC Storage with the new Dashboard from RAID Inc.

In this video from SC16, Yugendra Guvvala, VP of Technology at RAID Inc. describes the company’s new Dashboard software. The Dashboard provides a single pane of glass to manage your high performance Lustre storage pools. “Scaling to tens of petabytes and thousands of clients – considered a best filesystem for storage by many – Lustre is a high performance storage architecture for clusters. The central component of this architecture is the Lustre shared file system, which is currently available for Linux, providing a POSIX-compliant UNIX file system interface. RAID, Inc. offers custom Lustre solutions with installation & 24/7 support.”

RAID Inc. & SwiftStack Deliver HPC Object Storage at Scale

Today RAID Inc announced a strategic go to market partnership with SwiftStack, the leader in object storage for the enterprise. This officially ratified partnership augments the RAID Inc. portfolio of technical computing and data center infrastructure solutions by adding SwiftStack’s object storage platform to its Ability Storage solutions, ultimately allowing a lower total cost of ownership (TCO) for all. “State of the art advancements in computing require a storage architecture that traditional SAN or NAS infrastructure was never designed for,” said Don Jaworski, CEO of SwiftStack. “Our partnership with RAID Inc. gives HPC audiences a modern storage architecture that supports data-centric workflows and enables an infrastructure cost model that meets both the durability and budget requirements of those organizations.”