HPCwire: Intel and Cray Bring Collaboration Back to Supercomputing

The Leading Source for Global News and Information Covering the Ecosystem of High Productivity Computing

Leading HPC
Solution Providers

HPCwire >> Topic >> Processors

April 28, 2008

Intel and Cray Bring Collaboration Back to Supercomputing

-- You've Got the Brawn; I've Got the Brains

by Addison Snell, VP/GM Tabor Research

Intel and Cray have announced a partnership to begin collaboration on the design of a new supercomputer, which could be available in approximately three years. Under the partnership agreement, engineers from each company will work together directly, with the goal of generating a scalable architecture that will take unique advantage of next-generation Intel Xeon processors.

This is a significant announcement that has major potential benefits for both Intel and Cray. For Intel, the collaboration validates the Xeon line for high-end supercomputing, as Intel sets its sights on TOP500 domination. For Cray, Intel's partnership provides market velocity and support for the company's future Adaptive Supercomputing vision with a potential vehicle for downstream machines.

The joint development of a new supercomputer will also benefit users in the topmost echelon of the High Productivity Computing market. The challenge of efficient cross-system memory communications is compounding with the transition to multi-core, and both processor vendors and system vendors are reaching for new solutions. Collaboration between processor and system fabric architects gives the combined team the opportunity to exploit a full capability set and therefore approach the challenge from the standpoint of integrated system design, as opposed to distinct technologies that must be somehow fit together.

You've got the brawn / I've got the brains / Let's make lots of money
– Pet Shop Boys, "Opportunities"

In their 1986 U.S. top-ten hit "Opportunities," the U.K. electropop duo Pet Shop Boys synthesized a collaborative vision that could have foreshadowed the dilemma facing the supercomputing industry two decades later, if only MGM had had the foresight to feature the song in a sequel to WarGames in which the DOD struggles to upgrade the WOPR.

Back then supercomputers were self-contained devices, and one company controlled the development of the processors, the operating system, and the system interconnect. Today, technological decentralization, brought about by the cost-saving economies of scale of commoditization, has frequently been cited as a barrier to efficient scalability.

Cray and Intel realize that they each control an important piece of the overall puzzle. (In Cray's case, it's arguably two pieces: the scalable architecture and the operating environment.) However, when considering the two conspirators, deciding which is the "brawn" and which is the "brains" is a matter of metaphorical perspective.

In a literal sense, Intel provides the brains of the supercomputer: future-generation many-core Xeon processors with the promise of compounded increases in theoretical peak performance. Cray therefore provides the brawn: the combined muscle of the SeaStar system fabric and the Cray Linux environment (not to mention Cray excellence in packaging and facilities issues at scale and the power of the Cray brand).

Upon reflection, we can easily view the metaphor the other way around. The brains at Cray are masters of scalability, but they have lacked market force in recent years. Intel's substantial market power, and resources, might be exactly the brawn Cray needs to get back on track.

I can program a computer / Choose the perfect time
If you've got the inclination / I have got the crime

For Cray, Intel's inclination to grab its own top-ten hits among the world's supercomputing rankings is fortuitous. Cray has continued to put out new supercomputers based on AMD Opteron processors (viz. the Cray XT5 and XT5h from last year), and yet the company still lags significantly behind IBM in supercomputing market share and continues to face negative perceptions regarding viability. In addition, pockets of traditionalist Cray users have lamented that, more or less, they don't make 'em like they used to.

No offense is intended to AMD. Opteron has made substantial inroads in HPC, and Cray is not planning any changes to its current AMD-based product line. The reason Cray's users might celebrate the Intel collaboration probably has more to do with collaboration than with Intel. No one disputes the raw power of commodity processors. The question is, what could the architecture do with them if the companies worked not as supplier and OEM, but as development partners? Cray and Intel intend to find out.

I'm looking for a partner / Regardless of expense
Think about it seriously / You know it makes sense

Intel also stands to benefit substantially from the engagement. Although Intel dominates AMD in market share across every other computing category, the question of which processors power the world's most powerful supercomputers is presently unclear. AMD got a jump on Intel by being the first to offer technologies such as HyperTransport and on-chip memory controllers, both of which better allowed performance to scale. In recent studies, Tabor Research found Opteron to be offered as an option on as many HPC products as Xeon was, and our survey of supercomputing sites found more Opteron nodes than Xeon nodes. AMD's OEM arrangement with Cray -- a brand synonymous with supercomputing -- is an additional fly in Intel's ointment.

Intel has been making a comeback of late, no doubt aided by delays in the release of AMD's quad-core offering. However, we do not expect Intel to settle for trading cyclical leads with AMD, based on release schedules. Intel's partnership with Cray could be an instrument to capturing dominance in the supercomputing space and across the broader HPC market. If successful, the alliance puts a blemish on what luster AMD has achieved in HPC, which could result in substantial material benefits for Intel.

Intel and Cray did not disclose what agreement there was, if any, to exchange material wealth along with engineering ideas, leaving the final verse of analysis as a matter of conjecture. Tabor Research believes that at a minimum, Intel will provide processors to Cray for testing, development, and eventual delivery, at substantially discounted costs. Furthermore, we believe it is likely that Intel will cover certain development costs through "non-recurring expenditure" (NRE) payments to Cray. These payments, if made, would not only support development, but also provide Cray with the funding that would help ensure Intel had as strong a partner as possible when the release timeframe comes.

Oh, there's a lot of opportunities / If you know when to take them
There's a lot of opportunities / If there aren't, you can make them

This collaboration represents a tremendous opportunity for Intel and Cray, with the potential for both companies to "make lots of money." Beyond that, it is a partnership that bodes well for the industry, as it brings us back in time to an era when system development included all aspects of a system working together in harmony. Ultimately, the success of this partnership may hinge on each party's willingness to follow through. Everything is currently based on promises, and HPC has seen collaborations fail in the past, especially if they were inherently one-sided. If each company shares its know-how with the other, they have a remarkable opportunity to create a brains-plus-brawn supercomputer that benefits Intel, Cray, and the industry at large.

Article Tools

Share & Save Options

Discussion

There are 0 discussion items posted.

Join the Discussion

You must be a registered user in order to comment on HPCwire stories.

Members enjoy:

	Access to Special Content, including Podcasts, Webcasts, and Whitepapers and more!
	Ability to take an active role in the discussions that are shaping the HPC technology discipline!
	Special offers available only to HPCwire members!

Become a Registered User Today!

Registered Users Log in Here to Comment

Email Address:	Password (case sensitive)

Remember me	Forgot Password?

RSS Feeds

Feeds by Topic

Feeds by Industry

Feeds by Content Type

Feature Articles

The Week in Review

UPenn adds third state to nanowire storage; and UIUC is named the first CUDA Center of Excellence. John West recaps those stories and more in our weekly wrap-up.
Read More...

IBM Looks to Tap Massive Data Streams

Modern civilization is positively drenched in data, some of which needs to be dealt with in real time to be of any value. Businesses, especially in the financial industry, have long recognized this, and have been building custom systems to collect, analyze, and react to information as it is captured. IBM thinks the time is right to generalize these approaches into a new field of computing -- and a new business -- it calls stream computing.
Read More...

Gravity Attracts a GigE HPC Cluster

Not all supercomputing rides on InfiniBand or proprietary interconnects. For technical applications that decompose neatly into loosely-coupled threads, a big cluster with vanilla Gigabit Ethernet does just fine. The top Ethernet system on the TOP500 list -- at number 58 -- is the new ATLAS cluster at the Max Planck Institute for Gravitational Physics in Germany.
Read More...

Read more HPCwire features...

Top Headlines

San Diego Gets Set for Storage Explosion

Jul 03 | Byte and Switch | The San Diego Supercomputer Center, which provides much of the core storage for the TeraGrid, is overhauling its 28 petabyte storage system to support tremendous data growth. Read more...

Intel's Gelsinger Predicts Intel Inside Everything

Jul 03 | ExtremeTech | Intel exec Pat Gelsinger said he sees the Intel Architecture permeating virtually every segment of computing, as the company's microprocessors expand into more and more cores. Read more...

A Massively Parallel Future

Jul 03 | Bangkok Post | The latest programmable GPUs are starting to steal application cycles from CPUs. Read more...

UCSD Researchers Identify Potential Bird Flu Drugs

Jul 02 | UC San Diego News Center | With the help of resources at the San Diego Supercomputer Center, UCSD scientists have isolated more than two dozen promising compounds from which new “designer drugs” might be developed to combat the avian flu virus. Read more...

Implementing Multi-Core: The Devil Is in the Detail

Jul 02 | Chip Design Magazine | Dual- and quad-core processors barely scratch the surface of the potential of multi-core systems. Read more...

Read more headlines...

Featured Whitepapers

New HPC White Paper: Star-P® Performance on IBM Linux Clusters

Jul 03 | | The paper explores some of the performance benefits of Star-P on commodity scalable systems such as IBM's Linux clusters based on multi-core Intel Xeon processors. The results demonstrate substantial performance gains with almost no programmer effort-roughly a 24-fold speed improvement for solving linear matrix equations. An overview of parallel computing with Star-P, a description of the performance test cases and description of IBM cluster configurations used for testing are also addressed.

Fast N-Body Simulation with CUDA C Compiler

Apr 17 | | An N-body simulation numerically approximates the evolution of a system of bodies in which each body continuously interacts with every other body, and it arises in many other computational science problems as well.

Improving Performance and Manageability for Seismic Processing and Imaging Applications with Parallel Storage

Jun 05 | | As pressure increases on the upstream seismic processing community to deliver ever-higher levels of productivity and efficiency, a new generation of storage solutions will be required that allow the maximum utilisation of high-performance computing (HPC) Linux cluster resources, together with the minimum of management overhead.

View the White Paper Library

Multimedia

Podcast: Interview with Ben Bennett of ClearSpeed Technology

Today, HPC organizations are requiring substantially more floating point performance to solve real-world problems. In this podcast, Ben Bennett, ClearSpeed General Manager, discusses how acceleration technology can improve the overall performance of standard x86-based systems...