IBM, EMC win big Applera HPC server storage deal

Swinging Blue Genes

  • alert
  • submit to reddit

Internet Security Threat Report 2014

The Register has a new Enterprise Storage Channel

ComputerWire: IT Industry Intelligence

IBM Corp and EMC Corp were yesterday awarded contracts with Applera Corporation, the parent company of Celera Genomics Group which is best known as the cracker of the human genetic code,

Timothy Prickett Morgan writes

. Celera Genomics has been a long-time user of AlphaServer server clusters from the former Compaq Computer Corp, now part of Hewlett Packard Co, and Symmetrix storage arrays from EMC.

IBM is understandably ebullient about the 2 teraflops supercomputer based on clusters of its pSeries 690 Regatta-H Power4-based servers, which Applera will be installing to support Celera's computing needs as well as those of two other units of the company. The IBM servers will be replacing the installed base of AlphaServers at Applera's units, but the machines fall far short of the 100 teraflops monster supercomputer that Compaq and Celera Genomics were going to build with the help of Sandia National Labs, which is administered by the U.S. Department of Energy.

Part of the reason that Applera can get by with a smaller supercomputer - if one can call a 2 teraflops supercomputer small - is that it has in the past year been transforming itself from being a genome-cracker to a drug developer, with the drugs created based on its knowledge of the human genome and on its substantial understanding of how to use computers for molecular modeling.

Earlier this summer, Celera quietly backed out of the Compaq-Sandia partnership, and according to a Celera spokesperson, this was done because of changes in Celera's business model and because of changes in the Compaq product lines following the merger of Compaq and HP. Celera did not elaborate further, but it seems clear that Celera felt that if HP was going to push AlphaServer-Tru64 customers to HP-UX in the long term, given that its business model was changing anyway, it was a good time to open up the bidding for a new HPC cluster.

At about the same time this was going on, Sandia inked a $90m deal with Cray, Inc, the Seattle-based supercomputer maker under its "Red Storm" 100 teraflops supercomputer contract, which is being funded by the DOE and which had originally placed their order with Compaq in association with Celera back in January 2001. In July 2002, when Cray nosed in on the Red Storm contract from Sandia, we were given the impression that this did not necessarily spell doom for HP in pursuit of the Red Storm contract.

Sources at Sandia say that while the final negotiations for the Red Storm contract between Sandia and Cray are not done, an announcement awarding the deal to Cray is impending within the next few weeks and HP is not going to be awarded any portion of the Red Storm contract. Neither are IBM Corp, Sun Microsystems, SGI, by the way. Unless something goes terribly wrong, this will be an all Cray deal.

HP has the changes at Celera, the changes forced upon Compaq to kill off the Alpha processors in favor of Itanium chips, and the increasing aggressiveness of Cray to thank for its being bumped out of the Red Storm contract. This is a big deal, both in terms of money and prestige.

Red Storm calls for the 100 teraflops machine to be operational by 2004 and could account for hundreds of millions of dollars in hardware, software, storage, and services over the long haul. That said, HP is still the dominant supplier of HPC equipment, because of its strength in the midrange HPC market, even if other vendors may get the glory - however fleeting - of building the world's fastest supercomputer. HP takes down plenty of deals.

For instance, HP just closed an order for a 9.2 teraflops supercomputer for the DOE's Pacific Northwest national Laboratory that will be based on 1,400 future Itanium processors, a machine that is worth $24.5m. HP also this week announced a $22m deal at the Wellcome Trust Sanger Institute in the U.K. for a $22m AlphaServer cluster for genomics research that is comprised of 38 ES45s and four ES40 four-way AlphaServers, two DS20 two-way servers, and a single 32-way GS320 server, including three times the disk capacity of the original configuration.

IBM, of course, is no slouch when it comes to HPC. Machines like the Power4-based Regattas are designed specifically to take market share away from HP-Compaq in the HPC market, and it is not surprising that IBM could win the Applera deal and boot AlphaServer equipment from the account, regardless of the historical use of Compaq gear at Applera's main IT resource consumer, Celera Genomics.

IBM has shown a willingness to compete aggressively on price to get into new and high-profile accounts, and it undoubtedly did so at Applera. IBM says that the dozen 32-way pSeries 690 servers alone (which have been configured with only 64GB of memory apiece) have a list price of $25m, and the 150TB of disk capacity that Applera is acquiring from EMC is probably worth somewhere in the neighborhood of between $15m and $20m at list price. Considering the size of the deal, the poor economic conditions, and the desire to retain account control for EMC (you can bet IBM bid its Shark arrays as part of the deal) and to get into the account for IBM, it seems reasonable that the core hardware was acquired at around 50% of list price.

Applera did not say what software it was installing on the machine, but it is paying IBM for migration services and a multi-year services contract that will bring big bucks to IBM. The Regatta machines will run the forthcoming AIX 5.2 operating system (due to be announced in a few weeks) and will be clustered using IBM's latest "Colony" SP switching technology. This Regatta cluster will provide data processing support for Applera's Celera Genomics, Applied Biosystems (which sells the human genome data and other products and raked in $1.6bn in sales in 2001), and Celera Diagnostics (which makes medical equipment) units. The IBM machines will be located in Celera's Rockville, Maryland offices.

IBM also said that Applera and Big Blue are in preliminary discussions to develop life sciences products, although Peter Ungara, vice president of HPC at IBM, was not able to say what these products would be or who would sell them because discussions are only now getting underway.

© ComputerWire

The Register Storage Channel

Internet Security Threat Report 2014

More from The Register

next story
Docker's app containers are coming to Windows Server, says Microsoft
MS chases app deployment speeds already enjoyed by Linux devs
IBM storage revenues sink: 'We are disappointed,' says CEO
Time to put the storage biz up for sale?
'Hmm, why CAN'T I run a water pipe through that rack of media servers?'
Leaving Las Vegas for Armenia kludging and Dubai dune bashing
'Urika': Cray unveils new 1,500-core big data crunching monster
6TB of DRAM, 38TB of SSD flash and 120TB of disk storage
Facebook slurps 'paste sites' for STOLEN passwords, sprinkles on hash and salt
Zuck's ad empire DOESN'T see details in plain text. Phew!
SDI wars: WTF is software defined infrastructure?
This time we play for ALL the marbles
Windows 10: Forget Cloudobile, put Security and Privacy First
But - dammit - It would be insane to say 'don't collect, because NSA'
prev story


Forging a new future with identity relationship management
Learn about ForgeRock's next generation IRM platform and how it is designed to empower CEOS's and enterprises to engage with consumers.
Cloud and hybrid-cloud data protection for VMware
Learn how quick and easy it is to configure backups and perform restores for VMware environments.
Three 1TB solid state scorchers up for grabs
Big SSDs can be expensive but think big and think free because you could be the lucky winner of one of three 1TB Samsung SSD 840 EVO drives that we’re giving away worth over £300 apiece.
Reg Reader Research: SaaS based Email and Office Productivity Tools
Read this Reg reader report which provides advice and guidance for SMBs towards the use of SaaS based email and Office productivity tools.
Security for virtualized datacentres
Legacy security solutions are inefficient due to the architectural differences between physical and virtual environments.