Sign up for our newsletter and get the latest HPC news and analysis.


HPC News Roundup for Friday, May 8

hatnews

It can be tough to keep up with all that’s happening in HPC these days. Here are some notable items that may have not made it to the front page.

Lenovo Integrates Diablo’s Memory Channel Storage

diablo

Today Diablo Technologies announced that its Memory Channel Storage (MCS) architecture has been integrated into the latest edition of Lenovo’s X6 server solutions. The news comes on the heels of Diablo winning a series of legal battles against Netlist that have stalled deployment of the MCS technology.

Video: The Future of Interconnect with OpenPOWER

scott

“ConnectX-4 EDR 100Gb/s with CAPI support tightly integrates with the POWER CPU at the local bus level and provides faster access between the POWER CPU and the network device. We will discuss the latest interconnect advancements that maximize application performance and scalability on OpenPOWER architecture, including enhanced flexible connectivity with the latest Mellanox ConnectX-3 Pro Programmable Network Adapter.”

Purdue to Build Rice Supercomputer in a Single Day

pt-rice-computerLO

Today Purdue University announced that staff and volunteers will build it’s newest HPC cluster in a single day. The “Install Day” event will take place May 8.

Oakley Cluster Powers Satellite Surface Mapping for Disaster Relief in Nepal

Two university research teams are employing satellite imagery and supercomputers to produce high-resolution images to aid the Nepali earthquake relief effort. This image is a hillshade-rendered Digital Terrain Model image of the Kathmandu Valley, Nepal, created by SETSM software.

Researchers are using supercomputer automated surface mapping technology to help with disaster relief and longer-term stabilization planning efforts related to the recent earthquake in Nepal.

Supercomputing Black Hole Mergers with Blue Waters

Manuela Campanelli, Director of CCRG

“The mathematics involved in simulating these events is very sophisticated because one has to solve the equations of Einstein’s general relativity and magneto-hydrodynamics all together. The problem also requires very advanced supercomputers running programs on tens of thousands of CPUs simultaneously, and the use of sophisticated techniques for data extraction and visualization. Petascale numerical simulation is therefore the only tool available to accurately model these systems.”

New Intel Xeons Target Realtime Analytics

xeon

Today Intel announced the new Xeon processor E7-8800/4800 v3 product families, delivering accelerated business insight through real-time analytics.

BlueTides on Blue Waters: The First Galaxies

Figure 1 from The BlueTides Simulation paper; reproduced with permission; 
z = 8 refers to a redshift of 8 when the universe was a little over 1/2 billion years old.

“The largest high-redshift cosmological simulation of galaxy formation ever has been recently completed by a group of astrophysicists (Drs. Feng, Di-Matteo, Croft, Bird, and Battaglia) from the U.S. and the U.K. This tour-de-force simulation was performed on the Blue Waters Cray XE/XK system at NCSA and employed 648,000 cores. They utilized approximately 700 billion particles (!) to represent dark matter and ordinary matter and to create virtual galaxies inside the supercomputer. The authors, who represent Carnegie Mellon University, UC Berkeley, Princeton University, and the University of Sussex, have given their simulation the moniker BlueTides.”

Beowulf Boot Camp Returns to LSU July 6-10

2014Beowulf

Louisiana State University will host the Beowulf Boot Camp, July 6-10, on the LSU campus. The course offers students and teachers a unique opportunity to work with advanced research technology not usually available in a typical classroom setting.



Cyfronet Launches World’s Largest HP Apollo 8000 Supercomputer

Marek Magryś from Cyfronet describes the new Prometheus system.

Cyfronet recently celebrated the launch of Poland’s fastest supercomputer. As the world’s largest deployment of the HP Apollo 8000 platform, the 1.68 Petaflop Prometheus system is powered by 41,472 Intel Haswell cores and 216 Terabytes of memory.