With Windows Support, SGI Casts Altix UV in New Light

By Michael Feldman

April 3, 2011

SGI has been getting a lot of mileage out of its SGI UV shared memory platform, having delivered close to 500 systems since it started shipping them in June 2010. Now, with the recent addition of support for Microsoft’s Windows Server, the company is looking to expand its customer base in a big way.

Altix UV, SGI’s latest generation shared-memory supercomputer, was introduced at the Supercomputing Conference in November 2009. It uses SGI’s fifth generation NUMAlink interconnect technology and Intel “Nehalem” Xeon processors to construct HPC-class SMP server nodes. The interconnect, along with the special UV hub chip, glue all the processors and memory together so that they can be operated as a monolithic resource. A fully tricked-out Altix UV 1000 will have 2,048 cores (4,096 threads via HyperThreading) and 16 TB of globally shared memory. A maximally configured machine represents 18.5 teraflops of peak performance.

Being able to command all that power within a single system image has a number of advantages, the main one being you can run standard (non-MPI) applications on a machine that for all intents and purposes behaves as an enormous PC with gobs of cores and memory at its disposal. And, by definition, such a system doesn’t require the complex set-up, software licensing, and maintenance of a distributed cluster platform — not an easy task as you approach the 1000-core realm.

Up until a few weeks ago, Altix UV came only with Linux, either Novell’s SUSE or Red Hat’s enterprise version. In early March, support was added for Microsoft Windows Server 2008 R2. The first iteration supported up to 128 cores and 1 TB of memory. On March 25, the company announced Windows Server was certified to the OS’s maximum reach: 256 cores and 2 TB of memory.

IBM and HP also have large shared memory x86-based servers with Windows Server support. But IBM’s X3950 and HP’s Proliant DL980 G7 top out at 96 and 64, respectively — well below the Windows Server limits. “Our engineering work finally brings Windows into true scalability,” says SGI CEO Mark Barrenechea.

On the other hand, Itanium-based platforms on Windows can scale to 128 cores. But with the new UV-Windows set-up, those customers (principally HP Integrity users) can now migrate their codes to SGI UV gear and achieve even greater scalability, at least on the core-count side. Itaniums still prevail in memory reach, being able to access up 128 TB.

Barrenechea says they’re targeting two major application areas with this system, the first being SGI’s traditional technical computing market. The top five application suites they expect will take advantage of the Windows-UV combo are ANSYS FLUENT, MATLAB, Mathematica, LS-Dyna, and Accelerys. These run the gamut from CFD and FEA, to computational chemistry and computational biology.

The idea here is to allow scientists to take their PC-based codes and easily slide them into these big memory UV machines with little if any porting work. In some cases, they won’t even need to perform a recompilation. A PC binary should be able to run unaltered on the Xeon-based machine (although maybe not optimally), and if the code was written correctly, will automagically take advantage of the larger memory. Of course, to utilize additional UV cores, the developer will have to parallelize the code via OpenMP threading or the equivalent.

But many of these applications are constrained only by available memory, (requiring just one to four threads to do their job). Since a typical PC isn’t going to have more than a few gigabytes of RAM, the data sizes are going to be rather limited when it comes a traditional HPC simulation code. Even a relatively modest-sized four-dimensional array of 1000 x 1000 x 1000 x 1000 byte-sized elements (for say a 3D object moving through time) will occupy an entire terabyte.

At the recent HPCC conference in Newport, Rhode Island, SGI CTO Dr. Eng Lim Goh demonstrated a simulation of the human heart developed at the University of Montreal. On a laptop, because of the limited memory, it could only be run with 60 million grid points. That delivered a rather poor resolution of the heart in action. Moving it to an Altix UV machine with 1.2 TB of memory, the model was expanded to 2 billion grid points, providing a much more realistic model.

At that scale, the simulation still took two weeks to compute a single heartbeat. Goh suggested that parallelizing the code to take advantage of the additional UV cores (768 in this case) might be able speed up the model to something close to real-time.

But big memory is not just for technical workloads. The second major application area for a Windows-capable Altix UV is on the enterprise side, in the realm of data-intensive applications. In particular, we’re talking about data warehousing, data mining, business intelligence and related types of tools. The driver behind these applications is Microsoft’s SQL Server, whose support was added in conjunction with the Windows Server OS.

This area represents a new market for SGI, although some of these customers have HPC leanings as well. In general, though, any informatics-type application that encapsulates terascale-sized structured databases is fair game for an Altix UV. The fact that many of these codes are developed in and for a Microsoft environment means there is now an easier path to greater scalability.

Barrenechea considers SGI’s entry into Microsoft’s software ecosystem a significant step for them. “Sure, we’ve supported Windows and certified it,” he says, ‘but it’s a new focus for the company.”

Of course, Linux will be the operating system of choice for most HPC users. And, in fact, Altix UV scalability is still better on that OS. Red Hat Enterprise Linux 6 reaches to 8 TB of memory, while SUSE Linux Enterprise Server 11 hits the full 16 TB. Conveniently, Linux also supports all 2,048 cores of a top-end UV, although it’s hard to imagine an SMP-based code scaled to that level.

It should be noted that the memory limit on the Altix UV is actually constrained by the current generation of Xeon chips, whose 44-bit addressing scheme maxes out at 16 TB. If your data outgrows that capacity, Intel’s next-generation “Sandy Bridge” Xeons will add a couple more bits to quadruple its memory reach to 64 TB. According to SGI’s Goh, the company plans to support the new chips in an upcoming version of the Altix UV, and already have one order for such a system.

Core counts on the next-generation Altix UV may rise as well, although the most acute demand will remain on the memory capacity side. In any case, one or more of the supported OS’s will likely be tweaked to support any new limits SGI comes up with in future UV hardware.

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!

UCSD, AIST Forge Tighter Alliance with AI-Focused MOU

January 18, 2018

The rich history of collaboration between UC San Diego and AIST in Japan is getting richer. The organizations entered into a five-year memorandum of understanding on January 10. The MOU represents the continuation of a 1 Read more…

By Tiffany Trader

New Blueprint for Converging HPC, Big Data

January 18, 2018

After five annual workshops on Big Data and Extreme-Scale Computing (BDEC), a group of international HPC heavyweights including Jack Dongarra (University of Tennessee), Satoshi Matsuoka (Tokyo Institute of Technology), Read more…

By John Russell

Researchers Measure Impact of ‘Meltdown’ and ‘Spectre’ Patches on HPC Workloads

January 17, 2018

Computer scientists from the Center for Computational Research, State University of New York (SUNY), University at Buffalo have examined the effect of Meltdown and Spectre security updates on the performance of popular H Read more…

By Tiffany Trader

HPE Extreme Performance Solutions

HPE and NREL Take Steps to Create a Sustainable, Energy-Efficient Data Center with an H2 Fuel Cell

As enterprises attempt to manage rising volumes of data, unplanned data center outages are becoming more common and more expensive. As the cost of downtime rises, enterprises lose out on productivity and valuable competitive advantage without access to their critical data. Read more…

Fostering Lustre Advancement Through Development and Contributions

January 17, 2018

Six months after organizational changes at Intel's High Performance Data (HPDD) division, most in the Lustre community have shed any initial apprehension around the potential changes that could affect or disrupt Lustre Read more…

By Carlos Aoki Thomaz

UCSD, AIST Forge Tighter Alliance with AI-Focused MOU

January 18, 2018

The rich history of collaboration between UC San Diego and AIST in Japan is getting richer. The organizations entered into a five-year memorandum of understandi Read more…

By Tiffany Trader

New Blueprint for Converging HPC, Big Data

January 18, 2018

After five annual workshops on Big Data and Extreme-Scale Computing (BDEC), a group of international HPC heavyweights including Jack Dongarra (University of Te Read more…

By John Russell

Researchers Measure Impact of ‘Meltdown’ and ‘Spectre’ Patches on HPC Workloads

January 17, 2018

Computer scientists from the Center for Computational Research, State University of New York (SUNY), University at Buffalo have examined the effect of Meltdown Read more…

By Tiffany Trader

Fostering Lustre Advancement Through Development and Contributions

January 17, 2018

Six months after organizational changes at Intel's High Performance Data (HPDD) division, most in the Lustre community have shed any initial apprehension aroun Read more…

By Carlos Aoki Thomaz

When the Chips Are Down

January 11, 2018

In the last article, "The High Stakes Semiconductor Game that Drives HPC Diversity," I alluded to the challenges facing the semiconductor industry and how that may impact the evolution of HPC systems over the next few years. I thought I’d lift the covers a little and look at some of the commercial challenges that impact the component technology we use in HPC. Read more…

By Dairsie Latimer

How Meltdown and Spectre Patches Will Affect HPC Workloads

January 10, 2018

There have been claims that the fixes for the Meltdown and Spectre security vulnerabilities, named the KPTI (aka KAISER) patches, are going to affect applicatio Read more…

By Rosemary Francis

Momentum Builds for US Exascale

January 9, 2018

2018 looks to be a great year for the U.S. exascale program. The last several months of 2017 revealed a number of important developments that help put the U.S. Read more…

By Alex R. Larzelere

ANL’s Rick Stevens on CANDLE, ARM, Quantum, and More

January 8, 2018

Late last year HPCwire caught up with Rick Stevens, associate laboratory director for computing, environment and life Sciences at Argonne National Laboratory, f Read more…

By John Russell

Inventor Claims to Have Solved Floating Point Error Problem

January 17, 2018

"The decades-old floating point error problem has been solved," proclaims a press release from inventor Alan Jorgensen. The computer scientist has filed for and Read more…

By Tiffany Trader

US Coalesces Plans for First Exascale Supercomputer: Aurora in 2021

September 27, 2017

At the Advanced Scientific Computing Advisory Committee (ASCAC) meeting, in Arlington, Va., yesterday (Sept. 26), it was revealed that the "Aurora" supercompute Read more…

By Tiffany Trader

Japan Unveils Quantum Neural Network

November 22, 2017

The U.S. and China are leading the race toward productive quantum computing, but it's early enough that ultimate leadership is still something of an open questi Read more…

By Tiffany Trader

AMD Showcases Growing Portfolio of EPYC and Radeon-based Systems at SC17

November 13, 2017

AMD’s charge back into HPC and the datacenter is on full display at SC17. Having launched the EPYC processor line in June along with its MI25 GPU the focus he Read more…

By John Russell

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

IBM Begins Power9 Rollout with Backing from DOE, Google

December 6, 2017

After over a year of buildup, IBM is unveiling its first Power9 system based on the same architecture as the Department of Energy CORAL supercomputers, Summit a Read more…

By Tiffany Trader

Fast Forward: Five HPC Predictions for 2018

December 21, 2017

What’s on your list of high (and low) lights for 2017? Volta 100’s arrival on the heels of the P100? Appearance, albeit late in the year, of IBM’s Power9? Read more…

By John Russell

GlobalFoundries Puts Wind in AMD’s Sails with 12nm FinFET

September 24, 2017

From its annual tech conference last week (Sept. 20), where GlobalFoundries welcomed more than 600 semiconductor professionals (reaching the Santa Clara venue Read more…

By Tiffany Trader

Leading Solution Providers

Chip Flaws ‘Meltdown’ and ‘Spectre’ Loom Large

January 4, 2018

The HPC and wider tech community have been abuzz this week over the discovery of critical design flaws that impact virtually all contemporary microprocessors. T Read more…

By Tiffany Trader

Perspective: What Really Happened at SC17?

November 22, 2017

SC is over. Now comes the myriad of follow-ups. Inboxes are filled with templated emails from vendors and other exhibitors hoping to win a place in the post-SC thinking of booth visitors. Attendees of tutorials, workshops and other technical sessions will be inundated with requests for feedback. Read more…

By Andrew Jones

Tensors Come of Age: Why the AI Revolution Will Help HPC

November 13, 2017

Thirty years ago, parallel computing was coming of age. A bitter battle began between stalwart vector computing supporters and advocates of various approaches to parallel computing. IBM skeptic Alan Karp, reacting to announcements of nCUBE’s 1024-microprocessor system and Thinking Machines’ 65,536-element array, made a public $100 wager that no one could get a parallel speedup of over 200 on real HPC workloads. Read more…

By John Gustafson & Lenore Mullin

Delays, Smoke, Records & Markets – A Candid Conversation with Cray CEO Peter Ungaro

October 5, 2017

Earlier this month, Tom Tabor, publisher of HPCwire and I had a very personal conversation with Cray CEO Peter Ungaro. Cray has been on something of a Cinderell Read more…

By Tiffany Trader & Tom Tabor

Flipping the Flops and Reading the Top500 Tea Leaves

November 13, 2017

The 50th edition of the Top500 list, the biannual publication of the world’s fastest supercomputers based on public Linpack benchmarking results, was released Read more…

By Tiffany Trader

GlobalFoundries, Ayar Labs Team Up to Commercialize Optical I/O

December 4, 2017

GlobalFoundries (GF) and Ayar Labs, a startup focused on using light, instead of electricity, to transfer data between chips, today announced they've entered in Read more…

By Tiffany Trader

How Meltdown and Spectre Patches Will Affect HPC Workloads

January 10, 2018

There have been claims that the fixes for the Meltdown and Spectre security vulnerabilities, named the KPTI (aka KAISER) patches, are going to affect applicatio Read more…

By Rosemary Francis

HPC Chips – A Veritable Smorgasbord?

October 10, 2017

For the first time since AMD's ill-fated launch of Bulldozer the answer to the question, 'Which CPU will be in my next HPC system?' doesn't have to be 'Whichever variety of Intel Xeon E5 they are selling when we procure'. Read more…

By Dairsie Latimer

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