Skip to main content
Premium Trial:

Request an Annual Quote

PNNL Buys 8.3-Teraflop HP Computer For Computational Biology Research


While news of the “preliminary” approval of Hewlett-Packard’s proposed Compaq merger may have overshadowed it, HP closed a significant deal last week in its bid to earn respect in the high-performance computing market for life sciences. The DOE’s Pacific Northwest National Laboratory agreed to purchase a $24.5 million HP Linux-based supercomputer for its Molecular Sciences Computing Facility.

HP has traditionally lagged behind competitors such as IBM and Cray in securing DOE accounts, but this purchase will give HP bragging rights as the builder of the world’s most powerful Linux-based supercomputer. HP has stepped up its efforts to gain greater visibility in the high-performance computing market, a fact borne out by its advance from fifth to second place in the most recent Top 500 supercomputing ranking [BioInform 11-19-01].

The PNNL purchase “clearly illustrates that there is more than one top player in the supercomputing market,” said HP’s CTO Rich DeMillo.

Acknowledging that HP systems are a relative rarity in DOE computer labs, Scott Studham, technical group leader of computer operations for the PNNL facility, said that several factors of the system “stood out” in the lab’s evaluation process. One deciding factor was the sustained performance of the system, which runs at 80 percent of peak power — a level not attainable with other systems, Studham said. In addition, he described the HP super computer as “well balanced.”

Delivery of the supercomputer will begin in mid-2002, and it is scheduled to be fully operational in 2003. The 1,400-processor system is expected to be more than 30 times faster, have 50 times more disk space, and have 10 times more memory than PNNL’s current computer, which was one of the world’s most powerful when installed in 1997.

The new computational capacity will be a boon for PNNL researchers, who intend to run environmental and chemical simulations as well as systems biology, genomics, and proteomics applications on the system. David Dixon, associate director of theory, modeling, and simulation at PNNL, said the new machine would contribute to the lab’s research under the DOE’s Genomes to Life project. While PNNL has until now put its supercomputers to work simulating environmental waste flows and computational chemistry, the Genomes to Life work will focus on the effects of the environment on cells at the molecular level, giving the new machine a number of new tasks with “strong biological overtones,” according to Dixon.

Focus areas will include research into the effects of cellular interaction with metals and radiation as well as a genomics-based understanding of microbes of interest for bioremediation. In particular, Dixon said the lab would use the new computer to study proteins, including protein-protein interactions, protein-membrane interactions, protein-DNA interactions, and protein complexes. “This new machine will enable us to study much larger complexes than people have previously been able to study,” he said. Cell-signaling processes and networks will be another application area for the new computer.

But PNNL won’t be the only lab to benefit from the new machine. The Molecular Science Computing Facility provides access to its high-performance computing center — as well as its software suite and graphics and visualization lab — to outside scientists through a competitive proposal process.

PNNL’s New HP Supercomputer

• 1,400 next-generation Intel Itanium “McKinley” and “Madison” processors.

• Total peak performance of more than 8.3 teraflops.

• 1.8 terabytes of memory and 170 terabytes of disk space.

• Each of the 700 nodes will have an independent connection to a 53-terabyte storage area network.

• An additional 117 terabytes of usable local storage will be divided among the nodes.

• Final production system to be delivered in 2003.

— BT

Filed under

The Scan

Researchers Compare WGS, Exome Sequencing-Based Mendelian Disease Diagnosis

Investigators find a diagnostic edge for whole-genome sequencing, while highlighting the cost advantages and improving diagnostic rate of exome sequencing in EJHG.

Researchers Retrace Key Mutations in Reassorted H1N1 Swine Flu Virus With Avian-Like Features

Mutations in the acidic polymerase-coding gene boost the pathogenicity and transmissibility of Eurasian avian-like H1N1 swine influenza viruses, a PNAS paper finds.

Genome Sequences Reveal Evolutionary History of South America's Canids

An analysis in PNAS of South American canid species' genomes offers a look at their evolutionary history, as well as their relationships and adaptations.

Lung Cancer Response to Checkpoint Inhibitors Reflected in Circulating Tumor DNA

In non-small cell lung cancer patients, researchers find in JCO Precision Oncology that survival benefits after immune checkpoint blockade coincide with a dip in ctDNA levels.