New Haswells Spark Open Season for HPC Systems

By Nicole Hemsoth

September 15, 2014

What do you get when you add more cores, improved memory bandwidth, and a notable uptick in performance out of a freshly-launched chip? One hell of a lot of announcements from a broad, hungry base of server makers. And to be fair, a lot of excitement from the HPC user community to boot.

While one can expect a sweep of server upgrades from the “big box” server makers, for a processor family like the new Xeon E5-2600 v3 series that that targets a range of HPC workloads, there are set to be a bevy of tailored options from HPC-oriented vendors. From the big system vendors in high performance computing to the workstation gearheads, the much-anticipated release of the new “Haswell” chips, several variants of which are aimed squarely at HPC, has unleashed a storm of new offerings.

Before we launch into some select systems set to sport the new Haswells, it’s worth mentioning that a number of machines have felt an early impact of the arrival, including Intel early ship customers like the Pawsey Supercomputing Centre, whose “Magnus” system just tipped into petascale territory following a final stage upgrade of the new E5-2600 v3s to complete its 35,000 core Cray XC30 masterpiece. Other centers, including Finland’s CSC with its “Sisu” system, are finding far-reaching upgrades with the new cores. The size of the system has almost quadrupled with the addition of the new Haswells, offering 5x the performance over the previous state of the machine, bringing it to a theoretical peak performance point of 1.7 petaflops to power Finland’s scientific computing goals.

Both of these publicized early ship customers of the new Xeons are for large Cray systems. The supercomputing vendor was among many HPC hardware partners who announced the availability of the E5-2600 v3 series into their lines. Cray will be adding the new Haswells to its air and liquid-cooled CS and XC family of supercomputers and now has some early users to showcase well in advance of the rapidly-approaching SC14 event in New Orleans.

Not to be outdone, SGI also announced the availability of the new processors on its HPC systems, with the additional news that they achieved SPEC performance world records for the new Haswells on their ICE X system using the 12-core, 2.6 GHz (3.5 GHz at max Turbo) E5-2690 v3 chips. They also pointed to their upgrade story with the new processors outfitted on an ICE X machine at Imperial College London, where users saw nearly double the performance following the fitting for the newest Xeons.

Supermicro followed suit with its building block approach to high performance servers. The company rolled out its X10 line with the new Haswell chips, emphasizing the new 1U/2U Ultra SuperServers that can hit the max 18 core, 1.5TB of DDR4 limits with room support the Xeon Phi.

“Combined with our new X10 TwinPro, MicroBlade and FatTwin systems as well as the industry’s widest range of server building blocks with Titanium level high efficiency power supplies, Supermicro is delivering total solutions to address the critical power, space and cost challenges facing today’s data driven businesses,” said Charles Liang, CEO and President of Supermicro, whom we spoke with in advance of the release about the company’s larger emphasis on HPC workloads.

Other HPC-oriented vendors, including Penguin Computing, have added to their offerings with the addition of the Open Computing Project designed “Tundra” cluster platform, which is now available in the new Haswell flavor. The Tundra platform, which was announced at SC13, is capable of supporting 108 servers in a rack by leveraging an OCP-inspired approach that lets the company insert three nodes in 36 “open units,” which means they can fit 90 nodes into a traditional 42U rack. With the original rack, this led to a total of 40 teraflops for a single rack, which while there are still some variables if you’re trying to math out the current performance on those figures, is still far greater with the new Haswells—almost double at highest estimation using the 18-core variant. For the speculators, I’ll just leave this chart right here…

HaswellComparison

The dense design is coupled with their IceBreaker storage an Arctica switches for HPC and enterprise environments, providing what Penguin CEO, Tom Coull, describes as “TCO advantages for current large-scale HPC deployments while designed to accommodate future exascale HPC components such as coprocessors and new fabrics.”

NEC announced the launch of two new models of its Express5800 Series sporting the new E5-2600 v3 series. According to NEC, the new 2-socket rack-mount servers, equipped with high speed DDR4 memory, “enable processing performance to be improved by up to 40% when compared to previous models in addition to the 12Gb/s SAS RAID controllers and hard drives, which have strengthened I/O performance.”

The list of server makers who are ready to ship the new Haswell-based machines is quite long. Other notable mentions include AMAX, which has integrated the latest Xeon processor E5-2600 v3 family into its ever-broadening array of offerings in the HPC and big data arenas as well as smaller companies that are becoming more familiar names in HPC, including Silicon Mechanics and Advanced Clustering with their Pinnacle server line.

It’s not just the system vendors who are finding an in to talk about performance improvements that are around the bend with the new launch. Mellanox, for instance, verified that their InfiniBand and Ethernet offerings have been well-optimized for the new v3 product family with the target of delivering 100Gb/s bandwidth speeds and end-to-end latencies in the 650ns range.

“Intel’s new platform provides end-users with even greater application performance provided they are connected with interconnects supporting 100Gb/s and above bandwidth speeds,” said Gilad Shainer, vice president of marketing at Mellanox Technologies. “Mellanox EDR 100Gb/s InfiniBand and 40/56Gb/s Ethernet interconnect solutions will ensure that IT managers achieve the highest performance and return-on-investment in their data center server and storage upgrades.”

The real test of performance, efficiency, and overall real-world application will prove itself on the upcoming TOP500 list, which will provide a fair number of systems that were early customers for Intel’s newest addition to the Xeon family.

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industy updates delivered to you every week!

AI Thought Leaders on Capitol Hill

July 14, 2018

On Thursday, July 12, the House Committee on Science, Space, and Technology heard from four academic and industry leaders – representatives from Berkeley Lab, Argonne Lab, GE Global Research and Carnegie Mellon University – on the opportunities springing from the intersection of machine learning and advanced-scale computing. Read more…

By Tiffany Trader

HPC Serves as a ‘Rosetta Stone’ for the Information Age

July 12, 2018

In an age defined and transformed by its data, several large-scale scientific instruments around the globe might be viewed as a ‘mother lode’ of precious data. With names seemingly created for a ‘techno-speak’ glossary, these interferometers, cyclotrons, sequencers, solenoids, satellite altimeters, and cryo-electron microscopes are churning out data in previously unthinkable and seemingly incomprehensible quantities -- billions, trillions and quadrillions of bits and bytes of electro-magnetic code. Read more…

By Warren Froelich

Can Markov Logic Take Machine Learning to the Next Level?

July 11, 2018

Advances in machine learning, including deep learning, have propelled artificial intelligence (AI) into the public conscience and forced executives to create new business plans based on data. However, the scarcity of hig Read more…

By Alex Woodie

HPE Extreme Performance Solutions

Introducing the First Integrated System Management Software for HPC Clusters from HPE

How do you manage your complex, growing cluster environments? Answer that big challenge with the new HPC cluster management solution: HPE Performance Cluster Manager. Read more…

IBM Accelerated Insights

ORNL Summit Supercomputer Is Officially Here

Oak Ridge National Laboratory (ORNL) together with IBM and Nvidia celebrated the official unveiling of the Department of Energy (DOE) Summit supercomputer today at an event presided over by DOE Secretary Rick Perry. Read more…

CSIR, Nvidia Partner to Launch GPU-Powered AI Center in India

July 10, 2018

As reported by a number of Indian news outlets, India’s Council of Scientific and Industrial Research (CSIR) is partnering with Nvidia to establish a new, AI-focused Centre of Excellence in New Delhi, India's capital. Read more…

By Oliver Peckham

AI Thought Leaders on Capitol Hill

July 14, 2018

On Thursday, July 12, the House Committee on Science, Space, and Technology heard from four academic and industry leaders – representatives from Berkeley Lab, Argonne Lab, GE Global Research and Carnegie Mellon University – on the opportunities springing from the intersection of machine learning and advanced-scale computing. Read more…

By Tiffany Trader

HPC Serves as a ‘Rosetta Stone’ for the Information Age

July 12, 2018

In an age defined and transformed by its data, several large-scale scientific instruments around the globe might be viewed as a ‘mother lode’ of precious data. With names seemingly created for a ‘techno-speak’ glossary, these interferometers, cyclotrons, sequencers, solenoids, satellite altimeters, and cryo-electron microscopes are churning out data in previously unthinkable and seemingly incomprehensible quantities -- billions, trillions and quadrillions of bits and bytes of electro-magnetic code. Read more…

By Warren Froelich

Tsinghua Powers Through ISC18 Field

July 10, 2018

Tsinghua University topped all other competitors at the ISC18 Student Cluster Competition with an overall score of 88.43 out of 100. This gives Tsinghua their s Read more…

By Dan Olds

HPE, EPFL Launch Blue Brain 5 Supercomputer

July 10, 2018

HPE and the Ecole Polytechnique Federale de Lausannne (EPFL) Blue Brain Project yesterday introduced Blue Brain 5, a new supercomputer built by HPE, which displ Read more…

By John Russell

Pumping New Life into HPC Clusters, the Case for Liquid Cooling

July 10, 2018

High Performance Computing (HPC) faces some daunting challenges in the coming years as traditional, industry-standard systems push the boundaries of data center Read more…

By Scott Tease

Meet the ISC18 Cluster Teams: Up Close & Personal

July 6, 2018

It’s time to meet your ISC18 Student Cluster Competition teams. While I was able to film them live at the ISC show, the trick was finding time to edit the vid Read more…

By Dan Olds

PRACEdays18 Keynote Allan Williams (Australia/NCI): We’re Open for Business Down Under!

July 5, 2018

The University of Ljubljana in Slovenia hosted the third annual EHPCSW18 and fifth annual PRACEdays18 events which opened with a plenary session on May 29, 2018 Read more…

By Elizabeth Leake (STEM-Trek for HPCwire)

HPC Under the Covers: Linpack, Exascale & the Top500

June 28, 2018

HPCers can get painted as a monolithic bunch by outsiders, but internecine disagreements abound over the HPCest of HPC jargon, as was evident at ISC this week. Read more…

By Tiffany Trader

Leading Solution Providers

SC17 Booth Video Tours Playlist

Altair @ SC17

Altair

AMD @ SC17

AMD

ASRock Rack @ SC17

ASRock Rack

CEJN @ SC17

CEJN

DDN Storage @ SC17

DDN Storage

Huawei @ SC17

Huawei

IBM @ SC17

IBM

IBM Power Systems @ SC17

IBM Power Systems

Intel @ SC17

Intel

Lenovo @ SC17

Lenovo

Mellanox Technologies @ SC17

Mellanox Technologies

Microsoft @ SC17

Microsoft

Penguin Computing @ SC17

Penguin Computing

Pure Storage @ SC17

Pure Storage

Supericro @ SC17

Supericro

Tyan @ SC17

Tyan

Univa @ SC17

Univa

  • arrow
  • Click Here for More Headlines
  • arrow
Do NOT follow this link or you will be banned from the site!
Share This