In what some industry watchers view as a big loss for HP, Celera Genomics Group and sister-company Applied Biosystems Group have switched away from their five-year-old alliance with Compaq, signing storage and supercomputing deals with deals with EMC and IBM, respectively.
Celera Genomics Group, the company that mapped the human genome, had previously had a strong relationship with Compaq prior to its merger with HP in May. Compaq’s technology helped the firm beat public researchers in completing the map of the genome, with the final assembly computations for the project conducted on Compaq’s AlphaServer GS160. Applera also had over 700 CPUs and 100 terabytes of storage from Compaq running its data center.
“Our previous computing systems, and the relationship that we established with Compaq had to do with genome sequencing and assembly, which was our primary business objective a couple of years ago,” said Jamie Lacey, a spokeswoman for Celera Genomics. “Now, it’s a couple of years later, and the platform is a couple of years old, and for Celera we have very new business needs, as does Applied Biosystems.”
Lacey refused to discuss the terms of the deal or what advantages led them to the decision, stating only: “We went with the vendors that had the best solutions to meet our current business needs for flexibility, cost and performance.”
HP wasn’t happy, but took it in stride.
“We’re really disappointed about Celeron moving to another vendor,” said Kathy Sowards, a spokesperson for HP’s enterprise business. “But they are only one of many life-sciences customers for Hewlett Packard.”
Applera, the parent company of both groups, is working on an upgrade to its computational IT infrastructure supporting Celera’s drug discovery and development efforts and the Knowledge Business of Applied Biosystems.
Through the deal inked with IBM, Applera will deploy a supercomputing infrastructure consisting of 12 IBM eServer p690 systems running AIX, IBM’s UNIX operating system. The system, when fully configured, is expected to have a peak capacity of two teraflops, or two trillion calculations per second.
IBM’s part of the deal, which the company notes is worth somewhere in the ballpark of $25 million, is certainly a boon for the company, although some have speculated that the profits from the deal may be less than impressive.
Peter Morrissey, worldwide life sciences executive for IBM, however, said that simply was not the case.
“When we looked at the opportunity to partner and engage with Applera, we looked from both a business perspective as well as what the impact of a business relationship was,” said Morrissey. “It is not the case that we did not look at profit on the deal inn order to gain the business.”
Despite the gains from the p690 systems, Big Blue missed out on the storage end of the deal, which was given instead to EMC. The company will provide both Celera and Applied Biosystems with an updated automated networked storage infrastructure, including 150 terabytes of networked storage, file servers and control software, to replace its present storage systems.
The Celera spokeswomen refused to address details of why they selected separate vendors for the two deals, noting only that after a long evaluation process of numerous vendors, the combination of the two separate companies best suited Celera and Applied Biosystems’ current needs.
Morrissey said that while the company was disappointed at not landing the storage deal, that it would not be the first time IBM has worked as part of a heterogeneous system.
“Did we want to win all the business? The answer is yes,” said Morrissey. “It is the case that IBM has engaged in many partnerships with customers where it is a very heterogeneous environment and we’re very comfortable with it.”
As part of the deal with IBM, the two Applera businesses are also expected to explore potential collaborations with Big Blue on an array of initiatives designed to accelerate drug discovery and development through the use of information technologies. These initiatives include joint research projects to solve computationally intensive problems related to identifying new drug targets, and developing and co-marketing new life sciences solutions based on both companies’ products and technologies.