ASPIRE receives $15.6M award from DARPA PERFECT program

Krste Asanovic

The ASPIRE project at UC Berkeley has received a $15.6M award over 5.5 years as part of the DARPA PERFECT program. The ASPIRE project will be investigating how to provide the utmost energy efficiency for future DoD high-performance embedded applications, with an integrated effort spanning applications, algorithms, programming systems, architectures, and resilient circuits.  The program has an ambitious goal of achieving over 75GLOPS/W for double-precision computations in an entire system, or approximately 50x greater energy efficiency than the best of today’s embedded systems.


For Whom the Bell Tolls: Big D v. Big C

David Patterson

By: Peggy Aycinena

For whom the bell tolls: Big D v. Big C

Randy Royce died yesterday of cancer. You probably didn’t know him. I didn’t know him very well, but his wife and I went to high school together. The first time I met Randy, Yvette was sitting on his lap at our big noisy 10-year class reunion. I always remembered that, because you never saw two people who had more fun together, who loved each other more.

Randy was mayor of San Carlos in recent years, and worked for many years before that at HP and Agilent. A long time ago, we were with Randy and Yvette at a dinner party given by mutual friends. Randy mentioned that he knew Paul Otellini from college and I said, “Really? I’d like to meet Paul Otellini. I’d liked to interview him.”

Randy said, “You’re a journalist? So who do you write for?”

I told him I’d send him my credentials and he said he’d pass them along to Otellini. Sometime later, I heard back from Randy: No go. Otellini didn’t know who I was and wasn’t willing to submit to an interview. Oh well, you can’t win them all.

There is one thing I wish we could win, however, and that’s this blasted battle with cancer. I’m really sick of this disease and I’m sick of going to funerals. I went to a funeral last month in Sacramento for a family member who died of cancer at 59, I expect to go to another funeral next month in San Francisco for somebody who’s only 45,  and that doesn’t include Randy’s funeral sometime in the next few weeks. He was only 62.

Sitting here, writing this while overlooking the escalators in the Santa Clara Convention Center as SNUG 2013 unfolds below me, I’m reminded of seeing my brother-in-law in this place in November 1999. We met up here for lunch some weeks after his wife, my sister, died of cancer. He was still a walking zombie, still living in a state of shock. My sister was only 51 when she died after a brief 6-month battle with cancer, and he didn’t know where to start to put his life back on track, how to finish raising their kids.

I hate cancer. Particularly because it seems like everybody who dies of it, is so not done. They’re not done being married, raising their kids, meeting their grandchildren, saving the world. They’ve still got lots of irons in the fire, and lots more potential for making a difference in a world that they’re exiting way too soon. That was certainly true for my sister, and it was most certainly true for Randy Royce.

Anyway, that’s the voice of despair talking. If you want to hear the voice of hope and optimism, you should listen instead to U.C. Berkeley’s Dave Patterson. True, Prof. Patterson’s had previous successes in other areas of technology, but if his talk at the February 14th Berkeley EECS Annual Research Symposium is any indication, his future success is what he hopes to be remembered for in the long run, because he’s got a really big iron in the fire.

Big D.Patterson wants to use Big D.Data to beat the Big C.Cancer.

He wants to use the same number crunching concepts – gathering, sorting, analyzing – to track Killer Cancers that other crunchers of Big Data use to track Killer Weather, to predict Killer Bugs on billion-transistor chips, and to design Killer Apps that benefit from Crowd Sourcing.

In other words, phenomenon which up-close are hard to understand, but looked at from way-far-away are characterized by zillions of data points that seem to show patterns of behavior that could possibly be used to predict and, in some cases, prevent that behavior.

Things like Cancer – a complex systemic problem, but also a point problem. Every cancer is different because every single person is different, but there are patterns there – at the molecular level, at the organelle level, at the cellular level, at the organ level, and at the organism level.

If manipulating Big D could give us a better handle on how to predict Big C, perhaps with time and determination it would allow us to prevent the disease in the first place, or at least stop its spread even after it’s got a foothold in the host. At least that’s how it’s supposed to work.

Of course, Prof. Patterson’s not the only researcher working on these ideas, but he’s the one I most recently heard talking about it. And he’s certainly one of the people who seems hellbent on making a difference by stopping the carnage that Big C is causing in the Big W  –  a Big World which is a sadder and less brilliant place for the loss of Randy Royce yesterday, and my sister 13 years ago.

I wish Big D.Patterson and his colleagues every success in their efforts.

For full article click  Here