NY sinks more money into HPC despite troubled past

Print Friendly, PDF & Email

The New York State Foundation for Science, Technology and Innovation (NYSTAR) announced yesterday that it has awarded a 3 year, $3M grant for the development of the High Performance Computation Consortium (HPC2).

The consortium will provide hands on assistance, through their scientists and engineers to the broader New York State user community, both to ease access to supercomputing resources at the institutions and to guide users in the utilization of HPC. This award will allow for the computational scientists hosted in one of three centers, to be deployed to any location in the state to provide assistance in the use of the HPC assets to businesses and academic users.

Consortium members include Rensselaer Polytechnic Institute, Stony Brook University, the University at Buffalo and the New York State Education and Research Network (NYSERNet).

NYSTAR is (as far as I can tell) publicly funded, and exists to help companies navigate the funding vacuum between technology prototype and productization. New York has a troubled recent past with publicly funded HPC investments. Early this year a controversy erupted over the $80M supercomputing investment the NYC school system made as the New York Post ran a story headlined “SCHOOLS COMPUTER AN $80M ‘DISASTER'” (all caps their choice, not mine).

I think it’s pretty clear that I’m a huge fan of HPC investments – I am a card-carrying member of the HPC faithful after all. But one has to be careful to avoid investing in an HPC infrastructure for any reason other than it being the right tool for a particular, specific job. They are simply too expensive to play “build it and they will come” with public money.

Hopefully NY has it right this time.

Comments

  1. I don’t think the $80Million investment in the “ARIS – Achievement Reporting and Innovation System” was an HPC investment. The system is a web based system for managing student records.

    Perhaps you are confusing that with the equally substantial investment in NY Blue, a large IBM Bluegene system hosted at Brookhaven National Lab and funded mainly by the state of NY.

    It would be a shame for people to get the impression that NY Blue was a disaster if in fact you are talking about the ARIS system.

  2. Indeed, Brian, I am talking about ARIS. The ARIS system was covered, both by IBM and the state of NY, as an HPC investment. The original announcement (at http://www.supercomputingonline.com/article.php?sid=13172) included the following text:

    “In order to measure how well public school students are learning, the New York City will spend $80 million on a supercomputer designed to analyze the performance of the city’s 1.1 million school kids. Mayor Bloomberg said the cost is worth it.”

    And the rest of that story, including other parts of the announcement and subsequent reporting, continue to tap on the HPC drum.

  3. I am familiar with the project, and I can assure you that ARIS has nothing to do with HPC. The article you refer to does not reference any source with respect to the project’s HPC nature, and your quote was from the article, not from anyone connected with ARIS or the NYC DOE.

    Brian Bonenfant accurately described ARIS’s nature. The HPC angle was picked up on by people who didn’t know what they were talking about, and repeated from there. Please don’t perpetuate the mistake.

  4. Don – fair enough. Thanks for the added insight.