SC13 in Review: The Week’s Top Newsmakers

By Nicole Hemsoth

November 28, 2013

This year at SC13 we counted over 100 HPC-specific announcements that hit the wires over the course of the week, many of which were from vendors, organizations and users that we were within a short walk across the show floor in Denver.

We wanted to point to some of the key newsmakers during the show, but before we delve deeper into some of their news items, there were some stories that deemed “best in show” during the course of the event. Our selections for “can’t miss” news items for SC13 include:

To get the full news firehose from last week, scroll down to the bottom of our special SC13 coverage page for a peek at top selected items.

Let’s take a more focused look at some of the week’s top newsmakers, beginning with the one company that seems to be everywhere (and they were hard to miss with their neon green scarves)…

NVIDIA

Before we jump directly into some of NVIDIA’s specific news, there are a few figures from the Top 500 BoF session that we’d like to share to highlight a few points.

nvidia_top_500nvidiashareperf

Rather hard trends to ignore–movements that are echoed by both Intersect360 Research and IDC, which noted during their breakfast event at SC13 that “way back” in 2011, only 28.2% of the sites they surveyed had adopted coprocessors or accelerators whereas their 2013 figures showed that a remarkable 76.9% of sites had swept up acceleration. While they note that a great deal of this is still in experimental phase, it’s nonetheless significant as it could mark the shape of the Top 500 in lists to come–and with some recent news from NVIDIA that we’ll get to in a moment–more enterprise datacenters.

General trends aside, there was plenty of specific news around NVIDIA during SC13–not to mention quite a bit of action at their booth, which hosted ongoing sessions and learning/engagement activities. News-wise the GPU giant had a strong showing with the official unveiling and detailing of its K40 accelerators, which provide a boost in both processing capacity and memory over their K20X.

While you can read more about the release in this in-depth feature from last week, suffice to say, a number of system vendors climbed on board with support, including Cray, AMAX, Supermicro, Boston Limited, Exxac and others. As we reported, the upgrade can mean significant performance improvements via the activation of more cores on the GPU and also through a new GPU Boost mode that lets the CUDA cores overclock.

Outside of this uptick in GPU capability, NVIDIA had other news that could be a boon to its future expansion in further environments. First, they announced that CUDA 6.0 will be available in 2014, which they said can push an 8x improvement to applications. The update includes some long-awaited features, most notably unified memory. CUDA 6.0 also now enables new drop-in libraries and “multi-GPU scaling” which lets re-designed BLAS and FFT GPU libraries scale their performance automatically across up to 8 GPUs per node, “delivering over nine teraflops of double precision performance per node, and supporting larger workloads than ever before (up to 512GB).” This feature can be used with their new BLAS drop-in library.

In another effort to extend the reach of GPU computing, this time into the enterprise datacenter, NVIDIA and IBM announced a partnership to collaborate on GPU-accelerated versions of several of IBM’s enterprise applications on Power systems. The companies noted that this is “the first time that GPU accelerator technology will move beyond the realm of supercomputing and into the heart of enterprise-scale datacenters.”

“This partnership will bring supercomputer performance to the corporate data center, expanding the use of GPU accelerators well beyond the traditional supercomputing and technical computing markets,” said Ian Buck, vice president of Accelerated Computing at NVIDIA. “It will also provide existing supercomputing and high performance computing customers with new choices and technologies to build powerful, energy-efficient systems that drive innovation and scientific discovery.”

Aside from these items and the notable momentum on the Top500, NVIDIA’s GPUs topped the list for green supercomputers and overall efficiency. All ten of the top placeholders on the Green500 (some excellent info on that list and its evolution from one of its founders, Kirk Cameron here) were powered by GPUs. Further, the Top500 BoF shared the following to echo the efficiency sentiment…

NVIDIA

 

Cray

As we move over to Cray, it’s worth pointing to their own involvement in NVIDIA’s news last week around the coming K40. The company was among the first to announce that it would support the updated GPU across its entire supercomputer line.

Given their history in HPC, it’s easy to make the claim that Cray is hard to ignore at SC, but this has really been their year to shine in some new ways. While their financials may not reflect knockout, wild growth, they’re working to innovate ahead of the curve, expanding into markets outside of scientific computing. In addition to the focus on their YarcData division, which is dedicated to delivering system and software solutions targeted at big data use cases, the company took another step in the enterprise direction with a new framework designed to allow Hadoop some easier hooks into their XC30 line of supers.

Cray was also behind the only new addition to this year’s top ten on the Top500—the 6.27 petaflop Piz Daint system at the Swiss National Supercomputing Centre—an XC30 powered by a Xeon E5/NVIDIA K20x duo and the Aries interconnect. Cray claimed a total of 48 entries on the Top500 this year, 18 of which were in the top 100 (and 2 in the top ten—Titan and now Piz Daint). Last November (SC12) they had a total of 55 systems, in 2011 they had 40, and in 2009 they had a 25 system share. In the years prior to 2009, they tended to hover around the 20 system share (give or take a few) regularly.

If the trend holds, they’ll be grabbing an ever-growing slice of Top500 pie, especially with the addition of some new systems and updates they’ve talked about that will be ripe for ISC’s list. For now, here is the vendor system share from this year:

CraySysShare

There were a few end user stories about new and build-out implementations at the show as well, including news that Cray was awarded a contract to expand its XC30 environment at the University of Stuttgart’s “Hornet” site. When the expansion is complete, Hornet will deliver peak performance of more than seven petaflops (quadrillion mathematical calculations per second) and 2.3 petabytes of additional Cray file system storage for Lustre (CLFS). Additionally, the Center for Computational Sciences (CCS) at the University of Tsukuba in Japan has pushed an NVIDIA and Intel-powered Cray CS300 cluster into production. The new system has been combined with the University’s current Cray super, and is providing researchers and scientists with 1.1 petaflops at their disposal.

On the programming front, Cray also announced another sweep across parts of its line with the latest release of the Cray Compiler Environment (CCE), which is available now on the CS300 machines. According to Cray, this move with the CCE “provides customers with a proven, familiar and HPC-optimized compiler for highly parallel environments.”

We’ve definitely been watching Cray this year—and so have you, according to our Reader’s Choice award votes. The company was handed ten separate HPCwire awards this year– three Readers’ Choice Awards and seven Editors’ Choice Awards. These together mark the most awards Cray has won in a single year. This also marks the tenth consecutive year Cray was selected for HPCwire awards.

CrayAwards

“We are truly honored by the scope of industry recognition for Cray this year,” said Peter Ungaro, president and CEO of Cray. “Credit is due to the hundreds of Cray employees who have committed their talents and energy to helping our customers solve the world’s most difficult computing challenges. As a company, we are intently focused on being a global leader in supercomputing, and the HPCwire awards are an acknowledgement of the results of working closely with our partners and our customers.”

Data Direct Networks

DDNThere seemed to be quite a bit of activity around the Data Direct Networks booth this year at SC as they held demos and showed off some their shiniest appliances against the din of the many other storage vendors that packed the floor.

At SC13 DDN rolled out two new models of its Storage Fusion Architecture(SFA) technology – the SFA12KX and the SFA12KXE. The company showcased the SFA12KX at SC13 and discussed the advanced processor technology and optimized OS, which they say to delivers up to 48GB/s and 1.4 Million IOPS from a single appliance.

The SFA12KXE leverages DDNTM In-Storage Processing technology (see image below) to back their own EXAScaler and GRIDScaler parallel file systems, as well as customer applications running natively within the storage array. The SFA12KXE is set to deliver up to 23GB/s of file system performance and eliminates external servers and storage networking to bring it into the “converged” camp that’s been garnering a great deal of attention.

DDN_inStorage

According to Jean-Luc Chatelain, Executive Vice President of Strategy and Technology at DDN, “DDN’s SFA12KX appliances are the foundation of many of the world’s most demanding and data-intensive environments, each requiring massive performance and scale without high cost. With features such as application-aware Flash caching, Real-Time I/O and in-storage processing, our SFX12KX appliances will perform at up to 48GB/s.”

DDN gave SC13 attendees something else to talk about last week beyond their SFA news via  their annual HPC Trends survey, which found that, perhaps unsurprisingly, storage I/O performance is a top priority for those designing HPC systems for “big data” workloads. The survey, which hit a cross-section of 60 of their end users, found that 68% of those polled agree that data and data storage has become the most strategic part of the HPC datacenter. As DDN noted, “Moreover, by a margin of two to one respondents, the survey also reveals that today’s storage technologies will need to undergo massive change to hit exascale proportions.” They also found that 78% “agree that hybrid storage is the evolutionary next step for HPC storage, combining both the performance and cost efficiency benefits required of storage at exascale.”

Data Direct Networks was another company that raked in a number of awards from HPCwire this year. DDN pulled in six separate awards for its work in manufacturing, government and industry collaborations, and financial services among other areas. The company also won Best HPC Storage Technology in conjunction with TACC as well as a Reader’s choice for Best Use of HPC in Oil and Gas for its role in BP’s new Center for High Performance Computing in Houston.

IBM

Outside of one major partnership news item, IBM only had a couple of light announcements at SC13, but before we dig into those specifics, just wanted to point to a trend that we’d noted before. Take a look below at the slide from this year’s Top 500 announcement that lists the top ten systems. Notice a trend, given that we’re talking now about Big Blue? (note: yellow highlight from BoF emphasis on the one new system–not related).

IBM_Top10

While this isn’t necessarily a surprise, it’s also worth noting a few other ways that IBM stole the listings last week at SC13. Take a look for instance at the following two charts, which show both the Green 500 supercomputer list and the Graph 500, which pits systems against one another based on their performance on the graph problem benchmark.

First, the Graph500

IBM_Graph500

And now the Green500

IBM_Green500

As we showed earlier, there is some disparity between the big players in the Top 500 and their smaller counterparts. While HP has an overall Top500 system share of 39% to IBM’s 33% (but with a performance share of 31.6% to HP’s 15.5%), it’s certainly worth pointing out that IBM is taking the cake in ways that few vendors can in the three big areas that matter—performance on LINPACK, performance on big data graph problems, and overall efficiency.

Back in the section on NVIDIA we highlighted a Top500 slide that pointed to the most efficient architectures and while IBM isn’t anywhere to be found there, NVIDIA and IBM’s partnership (again, highlighted earlier in the article) could turn that graphic around by the time next November rolls around with the POWER and GPU combination.

Outside of that announcement, IBM had a relatively quiet show but if there was one new item that slipped under the radar (at least in many of the conversations this attendee was around) this was certainly one. We should also point out that the Gordon Bell prize this year was awarded to researchers tapping Sequoia, providing IBM with another notch in its SC belt.

Other Strong Showings

We wanted to highlight a few other vendors that we watched at SC13 who had noteworthy and/or voluminous news. These include Mellanox, HP (which still holds the systems share on the Top 500) and Bull. Some of their key news items are linked below.

Mellanox

Mellanox InfiniBand Adoption Grows Across TOP500 Systems

NVIDIA RDMA Technology Supported on Mellanox InfiniBand Adapters

Mellanox InfiniBand Adapter Performance Revealed

Quantum Selects Mellanox as InfiniBand Partner

Mellanox Ranked 251 on Deloitte’s Technology Fast 500

Purdue Supercomputer Utilizes Mellanox InfiniBand Adapters

HP

TACC to Deploy Maverick for Visualization and Data Analysis

HP Servers Support TOP500 Supercomputers

Bull

University of Reims Supercomputer Ranked 5th on Green500 List

Bull Receives Honors in HPCwire Readers’ and Editor’s Choice Awards

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!

China’s Expanding Effort to Win in Microchips

July 27, 2017

The global battle for preeminence, or at least national independence, in semiconductor technology and manufacturing continues to heat up with Europe, China, Japan, and the U.S. all vying for sway. A fascinating article ( Read more…

By John Russell

Hyperion: Storage to Lead HPC Growth in 2016-2021

July 27, 2017

Global HPC external storage revenues will grow 7.8% over the 2016-2021 timeframe according to an updated forecast released by Hyperion Research this week. HPC server sales, by comparison, will grow a modest 5.8% to $14.8 Read more…

By John Russell

Exascale FY18 Budget – The Senate Provides Their Input

July 27, 2017

In the federal budgeting world, “regular order” is a meaningful term that is fondly remembered by members of both the Congress and the Executive Branch. Regular order is the established process whereby an Administrat Read more…

By Alex R. Larzelere

HPE Extreme Performance Solutions

HPE Servers Deliver High Performance Remote Visualization

Whether generating seismic simulations, locating new productive oil reservoirs, or constructing complex models of the earth’s subsurface, energy, oil, and gas (EO&G) is a highly data-driven industry. Read more…

India Plots Three-Phase Indigenous Supercomputing Strategy

July 26, 2017

Additional details on India's plans to stand up an indigenous supercomputer came to light earlier this week. As reported in the Indian press, the Rs 4,500-crore (~$675 million) supercomputing project, approved by the Ind Read more…

By Tiffany Trader

Exascale FY18 Budget – The Senate Provides Their Input

July 27, 2017

In the federal budgeting world, “regular order” is a meaningful term that is fondly remembered by members of both the Congress and the Executive Branch. Reg Read more…

By Alex R. Larzelere

India Plots Three-Phase Indigenous Supercomputing Strategy

July 26, 2017

Additional details on India's plans to stand up an indigenous supercomputer came to light earlier this week. As reported in the Indian press, the Rs 4,500-crore Read more…

By Tiffany Trader

Tuning InfiniBand Interconnects Using Congestion Control

July 26, 2017

InfiniBand is among the most common and well-known cluster interconnect technologies. However, the complexities of an InfiniBand (IB) network can frustrate the Read more…

By Adam Dorsey

NSF Project Sets Up First Machine Learning Cyberinfrastructure – CHASE-CI

July 25, 2017

Earlier this month, the National Science Foundation issued a $1 million grant to Larry Smarr, director of Calit2, and a group of his colleagues to create a comm Read more…

By John Russell

Graphcore Readies Launch of 16nm Colossus-IPU Chip

July 20, 2017

A second $30 million funding round for U.K. AI chip developer Graphcore sets up the company to go to market with its “intelligent processing unit” (IPU) in Read more…

By Tiffany Trader

Fujitsu Continues HPC, AI Push

July 19, 2017

Summer is well under way, but the so-called summertime slowdown, linked with hot temperatures and longer vacations, does not seem to have impacted Fujitsu's out Read more…

By Tiffany Trader

Researchers Use DNA to Store and Retrieve Digital Movie

July 18, 2017

From abacus to pencil and paper to semiconductor chips, the technology of computing has always been an ever-changing target. The human brain is probably the com Read more…

By John Russell

The Exascale FY18 Budget – The Next Step

July 17, 2017

On July 12, 2017, the U.S. federal budget for its Exascale Computing Initiative (ECI) took its next step forward. On that day, the full Appropriations Committee Read more…

By Alex R. Larzelere

Google Pulls Back the Covers on Its First Machine Learning Chip

April 6, 2017

This week Google released a report detailing the design and performance characteristics of the Tensor Processing Unit (TPU), its custom ASIC for the inference Read more…

By Tiffany Trader

Nvidia Responds to Google TPU Benchmarking

April 10, 2017

Nvidia highlights strengths of its newest GPU silicon in response to Google's report on the performance and energy advantages of its custom tensor processor. Read more…

By Tiffany Trader

Quantum Bits: D-Wave and VW; Google Quantum Lab; IBM Expands Access

March 21, 2017

For a technology that’s usually characterized as far off and in a distant galaxy, quantum computing has been steadily picking up steam. Just how close real-wo Read more…

By John Russell

HPC Compiler Company PathScale Seeks Life Raft

March 23, 2017

HPCwire has learned that HPC compiler company PathScale has fallen on difficult times and is asking the community for help or actively seeking a buyer for its a Read more…

By Tiffany Trader

Trump Budget Targets NIH, DOE, and EPA; No Mention of NSF

March 16, 2017

President Trump’s proposed U.S. fiscal 2018 budget issued today sharply cuts science spending while bolstering military spending as he promised during the cam Read more…

By John Russell

CPU-based Visualization Positions for Exascale Supercomputing

March 16, 2017

In this contributed perspective piece, Intel’s Jim Jeffers makes the case that CPU-based visualization is now widely adopted and as such is no longer a contrarian view, but is rather an exascale requirement. Read more…

By Jim Jeffers, Principal Engineer and Engineering Leader, Intel

Nvidia’s Mammoth Volta GPU Aims High for AI, HPC

May 10, 2017

At Nvidia's GPU Technology Conference (GTC17) in San Jose, Calif., this morning, CEO Jensen Huang announced the company's much-anticipated Volta architecture a Read more…

By Tiffany Trader

How ‘Knights Mill’ Gets Its Deep Learning Flops

June 22, 2017

Intel, the subject of much speculation regarding the delayed, rewritten or potentially canceled “Aurora” contract (the Argonne Lab part of the CORAL “ Read more…

By Tiffany Trader

Leading Solution Providers

Facebook Open Sources Caffe2; Nvidia, Intel Rush to Optimize

April 18, 2017

From its F8 developer conference in San Jose, Calif., today, Facebook announced Caffe2, a new open-source, cross-platform framework for deep learning. Caffe2 is the successor to Caffe, the deep learning framework developed by Berkeley AI Research and community contributors. Read more…

By Tiffany Trader

Reinders: “AVX-512 May Be a Hidden Gem” in Intel Xeon Scalable Processors

June 29, 2017

Imagine if we could use vector processing on something other than just floating point problems.  Today, GPUs and CPUs work tirelessly to accelerate algorithms Read more…

By James Reinders

Russian Researchers Claim First Quantum-Safe Blockchain

May 25, 2017

The Russian Quantum Center today announced it has overcome the threat of quantum cryptography by creating the first quantum-safe blockchain, securing cryptocurrencies like Bitcoin, along with classified government communications and other sensitive digital transfers. Read more…

By Doug Black

MIT Mathematician Spins Up 220,000-Core Google Compute Cluster

April 21, 2017

On Thursday, Google announced that MIT math professor and computational number theorist Andrew V. Sutherland had set a record for the largest Google Compute Engine (GCE) job. Sutherland ran the massive mathematics workload on 220,000 GCE cores using preemptible virtual machine instances. Read more…

By Tiffany Trader

Google Debuts TPU v2 and will Add to Google Cloud

May 25, 2017

Not long after stirring attention in the deep learning/AI community by revealing the details of its Tensor Processing Unit (TPU), Google last week announced the Read more…

By John Russell

Groq This: New AI Chips to Give GPUs a Run for Deep Learning Money

April 24, 2017

CPUs and GPUs, move over. Thanks to recent revelations surrounding Google’s new Tensor Processing Unit (TPU), the computing world appears to be on the cusp of Read more…

By Alex Woodie

Six Exascale PathForward Vendors Selected; DoE Providing $258M

June 15, 2017

The much-anticipated PathForward awards for hardware R&D in support of the Exascale Computing Project were announced today with six vendors selected – AMD Read more…

By John Russell

Top500 Results: Latest List Trends and What’s in Store

June 19, 2017

Greetings from Frankfurt and the 2017 International Supercomputing Conference where the latest Top500 list has just been revealed. Although there were no major Read more…

By Tiffany Trader

  • arrow
  • Click Here for More Headlines
  • arrow
Share This