Tuesday, November 29, 2011
Saturday, November 19, 2011
Intel Shrinks Supercomputer Into the Palm of Your Hand
Intel's Technical Computing Group chief Raj Hazra holds a 1 teraflop Knights Corner chip.
Back in the late 1990s, Justin Rattner got a special sense of satisfaction every time he drove by a nondescript Intel building in Beaverton, Oregon. Inside, researchers from Intel and Sandia National Labs were assembling the ASCI Red supercomputer, the first computer capable of doing one trillion calculations per second.
That’s not bad for a chip that just a few years ago seemed to be a failure. Knights Corner was built from the ashes of Intel’s failed graphical processing unit (GPU), called Larrabee.
Nvidia has been making inroads with its Tesla processors.
When it goes online in January 2013, Stampede will have 10,000 times the processing power of ASCI Red
“It’s a reminder of how fast this industry moves,” he says. “I spent a lot of my life with a lot of my co-workers designing ASCI Red … to think that I can hold that in my hand now, it’s humbling.”.
How Many Neutrinos Does It Take to Screw Up Einstein?
Results from a second experiment uphold the observation that neutrinos are moving faster than the speed of light. The OPERA collaboration, which first reported the superluminal neutrinos in September, has rerun the experiment and detected 20 new neutrinos breaking Einstein’s theoretical limit.
The findings are heartening to anyone hoping to see a major physics revolution in their lifetime. But scientists, as ever, are being cautious, and it will take an independent replication of the results by another team to even begin convincing many of them.
But a great deal of scrutiny remains.
“I can now say that the probability of the result being correct has increased from 1 in a million to one in 100 thousand,” wrote physicist Philip Gibbs on the viXra log (though he stressed that those numbers were merely illustrative and not actual calculated values).
Tommaso Dorigo, a physicist at CERN, noted on his blog that there are still other possible sources of error. For instance, the OPERA collaboration’s clock might not have a fine enough resolution to determine exactly when the neutrinos arrived. “The measurement therefore is only a ‘partial’ confirmation of the earlier result: It is consistent with it, but could be just as wrong as the other,” he wrote.
Ultimately, the only thing that would convince many in the field is if another team upholds the findings in an independent experiment. Plunkett, co-spokesperson for the Main Injector Neutrino Oscillation Search (MINOS) experiment at Fermilab, says that his collaboration expects to have results checking the OPERA findings in the spring of 2012.
Image: OPERA collaboration
Subscribe to:
Posts (Atom)