Rensselaer Polytechnic gets its own Watson

Rensselaer Polytechnic gets its own Watson

Summary: The Rensselaer Polytechnic Institute in the US is the first university to receive IBM's Watson supercomputer. It plans to use it for research, development, and Jeopardy! re-runs.

TOPICS: Big Data, IBM, Education

"Watson, how do you pronounce the word 'Rensselaer'?"

OK, so IBM's Watson supercomputer--you know, the one that was featured on the game show Jeopardy! in a man versus machine moment--is not Apple's Siri personal assistant. But the Troy, New York-based Rensselaer Polytechnic Institute will receive the system all the same, thanks to a grant that allows the American private research university to pursue deeper research and development into big data, analytics, and cognitive computing. In return, Watson's brain gains more experience in the field.

(Credit: IBM)

Rensselaer's private Watson will have 15 terabytes of storage, more than its TV-star predecessor, and allow 20 people--faculty, graduate students, and even some lucky undergrads--to access the system at any given time. After all, if you're going to train the next generation of data scientists, you've got to give them the goods.

(If you're unfamiliar with IBM's Watson, it is a supercomputer like no other, arriving at decisions by processing spoken queries and information in a way that mimics human brain processes--as best we know them, anyway--instead of merely regurgitating information from a database. It is "machine learning" in the flesh, so to speak, and it's of great interest to various business sectors, from healthcare to financial services.)

Rensselaer has faculty on hand who specialize in artificial intelligence, big data, and web science; having a Watson down the hall is like being a kid in a (very expensive, very geeky) candy store, I imagine. And that's IBM's hope: that Rensselaer researchers will, in using Watson, broaden its dataset and sharpen its reasoning.

Which is the point of hanging out at a university, no?

IBM has worked with other US universities (Michigan State, Rochester, Cornell) on Watson-related programming, but it's never before turned a system over free of charge. It's part of a Shared University Research Award granted by IBM's research division, and includes hardware, software, and support services.

The two institutions have a shared history, of course--besides being regional neighbors (they are about two hours apart as the crow flies), several IBMers come from RPI's hallowed halls. Rensselaer was also part of a 2011 effort to develop open architecture for Watson.

Topics: Big Data, IBM, Education

Andrew Nusca

About Andrew Nusca

Andrew Nusca is a former writer-editor for ZDNet and contributor to CNET. During his tenure, he was the editor of SmartPlanet, ZDNet's sister site about innovation.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Reminds me of why I want to go there....

    Still waitin on the acceptance letter there though... Rensselaer was apparently inundated with applications this year so I won't know if I was accepted until February at some time... I'd love to get some hands-on training with a computer like Watson, though.
    • Good luck!

      Our fingers are crossed for you.
    • A very good school

      I enjoyed my 5 years there and learned a lot (1966-1971).
    • Very geeky place!

      My son is a freshman there - if you like the "geek" atmosphere, you'll fit right in and likely be happy. Good luck!
  • Cool

    Upstate NY rocks!
    Rabid Howler Monkey
  • Suspicious...

    "15 terabytes of storage" isn't all that much (my home 'poota has 14...) I'm thinking that 15 terabytes figure refers to the amount of memory (RAM).
    • I did, too.

      The original version of this story had "memory," per IBM's official word, but then they reached out and corrected it to storage.
  • RPI has come a long way

    Back in about 1957, I was an IBM Applied Science Rep and RPI was one of my accounts. Then they had an IBM 650 Magnetic Drum Data Processing Machine that used a 4" x 16" drum rotating at 12,500 RPM that could store 2,00 10-digit words. Instructions had two addresses, one for data and the other the location of the next instruction. Programming was done with the Symbolic Optimizing Assembly Program (SOAP) that would try to minimize the latency for instruction fetch if you assembled the inner loops first.