Sign up for our Technology Newsletter 
Do Supercomputers Lead To Downtime?
As supercomputers grow more powerful, they’ll also become more susceptible to failure, thanks to the increased amount of built-in componentry. A few researchers at the recent SC12 conference, held last week in Salt Lake City, offered possible solutions to this growing problem.
Today’s high-performance computing (HPC) systems can have 100,000 nodes or more — with each node built from multiple components of memory, processors, buses and other circuitry. Statistically speaking, all these components will fail at some point, and they halt operations when they do so, said David Fiala, a Ph.D student at the North Carolina State University, during a talk at SC12.
The problem is not a new one, of course. When Lawrence Livermore National Laboratory’s 600-node ASCI (Accelerated Strategic Computing Initiative) White supercomputer went online in 2001, it had a mean time between failures (MTBF) of only five hours, thanks in part to component failures. Later tuning efforts had improved ASCI White’s MTBF to 55 hours, Fiala said.
But as the number of supercomputer nodes grows, so will the problem. “Something has to be done about this. It will get worse as we move to exascale,” Fiala said, referring to how supercomputers of the next decade are expected to have 10 times the computational power that today’s models do.
Today’s techniques for dealing with system failure may not scale very well, Fiala said. He cited checkpointing, in which a running program is temporarily halted and its state is saved to disk. Should the program then crash, the system is able to restart the job from the last checkpoint.
The problem with checkpointing, according to Fiala, is that as the number of nodes grows, the amount of system overhead needed to do checkpointing grows as well — and grows at an exponential rate. On a 100,000-node supercomputer, for example, only about 35 percent of the activity will be involved in conducting work. The rest will be taken up by checkpointing and — should a system fail — recovery operations, Fiala estimated.
Because of all the additional hardware needed for exascale systems, which could be built from a million or more components, system reliability will have to be improved by 100 times in order to keep to the same MTBF that today’s supercomputers enjoy, Fiala said.
Fiala presented technology that he and fellow researchers developed that may help improve reliability. The technology addresses the problem of silent data corruption, when systems make undetected errors writing data to disk.
IBM Sued Over Disaster
IBM has been hit with a multimillion-dollar lawsuit by chemical products manufacturer Avantor Performance Materials, which alleges that IBM lied about the suitability of an SAP-based software package it sells in order to win Avantor’s business.
In 2010, Avantor decided to upgrade its ERP (enterprise resource planning) platform to SAP software, according to the lawsuit, filed Thursday in U.S. District Court for the District of New Jersey.
“Seizing upon Avantor’s decision — and fully aware that, given the competitive pressures of Avantor’s industry, and the specialized demands of its customers, Avantor could not tolerate any disruptions in customer service — IBM represented that IBM’s ‘Express Life Sciences Solution’ … was uniquely suited to Avantor’s business,” the lawsuit states. “The Express Solution is a proprietary IBM pre-packaged software solution that runs on an SAP platform.”
But Avantor discovered a different truth after signing on with IBM, finding that Express Life was “woefully unsuited” to its business and the implementation brought its operations to “a near standstill,” according to the suit.
IBM also violated its contract by staffing the project with “incompetent and reckless consultants” who made “numerous design, configuration and programming errors,” it states.
In addition, IBM “intentionally or recklessly failed” to tell Avantor about risks to the project and hurried towards a go-live date, the suit alleges.
“To conceal the System’s defects and functional gaps, IBM ignored the results of its own pre-go-live tests, conducted inadequate and truncated testing and instead recommended that Avantor proceed with the go-live as scheduled — even though Avantor had repeatedly emphasized to IBM that meeting a projected go-live date was far less important than having a fully functional System that would not disrupt Avantor’s ability to service its customers,” the suit states.
The resulting go-live, which occurred in May, “was a disaster,” with the system failing to process orders properly, losing some orders altogether, failing to generate need paperwork for U.S. Customs officials and directing “that dangerous chemicals be stored in inappropriate locations,” the suit states.
Avantor has suffered tens of millions of dollars in monetary damages, as well as taken a hit to its reputation among partners and customers, the suit states.
Is A Cyber Warfare On The Horizon?
F-secure’s chief security researcher, Mikko Hypponen, has warned that we are entering into a cyber warfare revolution, and that governments will soon attempt to outdo each other based on their computer weapons’ prowess.
The internet security expert said-any future crisis between technically advanced nations will involve cyber elements.
His comments came after hearing last week that China and the US have been engaging in “war games” simulations.
“I wasn’t expecting [war games] so soon,” Hypponen said.
“I’m surprised and I think it is a good move because everybody is worried about escalation. The way to fight unnecessary escalations is that you know more about how the perceived enemy would act if there would be an escalation. War games are exactly that.”
It was Hypponen’s observations on the war games which led him to remark that we must look at “the bigger picture”.
“We’ve seen a revolution in defence technology and in technology generally over the past 60 to 70 years and I believe we are right now seeing the beginning of the next revolution: a cyber warfare revolution, which is going to as big as the revolutions we’ve seen so far in technology becoming part of defence, and part of wars,” he added.
Hypponen also predicted that it won’t be long before the world sees its first cyber arms race, including cyber war rehearsals to prove how strong countries are and boasting about their cyber skills to make other countries pay attention.
“Like nuclear in the sixties, cyber attacks are a deterrent and deterrents only work if your perceived enemies know that you have it,” he said.
AMD To Slash A 10th Of Its Workforce
Comments Off on AMD To Slash A 10th Of Its Workforce
Advanced Micro Devices Inc revealed a plan on Thursday to save about $200 million of operating costs in 2012 by cutting 10 percent of its global workforce and streamlining internal business processes.
The layoffs mark the first major move by Chief Executive Rory Read, who took the helm in August to try to galvanize a microprocessor maker that has bled market share to larger rival Intel Corp, while missing out on the mobile device boom.
“It’s not too surprising given the operating background of the new CEO and this is exactly what you’d bring an outsider in to do, but their problems go far deeper right now,” said Alex Gauna, an analyst at JMP Securities.
The layoffs should be completed in 2012′s first quarter, AMD said in a statement. Savings generated could help bankroll research and expansion into areas such as low-power chips, emerging markets and cloud computing next year.
In late September AMD, a distant second to Intel in selling microprocessors that are the brains of PCs, warned of manufacturing problems manufacturing it new 32 nanometer Llano chips as well as older 45 nanometer chips.
Pentagon Practices Cyberwar
June 23, 2011 by admin
Filed under Around The Net
Comments Off on Pentagon Practices Cyberwar
A mock Internet where the Pentagon can practice cyberwar games — complete with software that simulates human behavior under multiple military threat levels — is due to be up and running in a year’s time, according to a published report.
Called the National Cyber Range, the computer network mimics the architecture of the Internet so military planners can study the effects of cyberweapons by acting out attack and defense scenarios, Reuters says.
Planning for the Cyber Range was carried out by Lockheed Martin, which won a $30.8 million Defense Advanced Research Projects Agency (DARPA) grant, and Johns Hopkins University Applied Physics Laboratory, which won $24.7 million.
Cyber Range plans call for the ability to simulate offensive and defensive measures of the caliber that nations might be able to carry out. DARPA wants the range to support multiple tests and scenarios at the same time and to ensure that they don’t interfere with each other. “The Range must be capable of operating from Unclassified to Top Secret/Special Compartmentalized Information/Special Access Program with multiple simultaneous tests operating at different security levels and compartments,” according to DARPA’s announcement of the project.
Google Facing DOJ Probe
May 13, 2011 by admin
Filed under Around The Net
Comments Off on Google Facing DOJ Probe
Google is under investigation by the U.S. Department of Justice concerning practices within its advertising program, and has set aside $500 million for a potential resolution fund, the company said in a regulatory filing on Tuesday.
In its quarterly report filed with the U.S. Securities and Exchange Commission, Google said that in May it accrued $500 million for the three-month period ended March 31, 2011 in connection “with a potential resolution of an investigation by the United States Department of Justice into the use of Google advertising by certain advertisers”.
Although it cannot predict the ultimate outcome of the matter, Google said that it believes that it will not have a material adverse effect on its business, consolidated financial position, results of operations, or cash flows.
The company did not explain why the charge had been taken retrospectively on its first quarter earnings. It is also not clear to whom Google would have to make the payment in the event of settlement.
Google declined to comment, stating that it was a legal matter. The company has updated its first-quarter results press release on its web site, to reflect the new charge.
Google Launches Online Magazine
March 27, 2011 by admin
Filed under Around The Net
Comments Off on Google Launches Online Magazine
Google has decided to launch its own quarterly online magazine, Think Quarterly, out of its centers in the U.K. and Ireland, saying that “in a world of accelerating change, we all need to take time to reflect.”
The first issue of Think Quarterly is already freely available online and is focuses on data, including data obesity, data impotence, data overload and open data.
“Think Quarterly is a unique communications tool that brings together some of the world’s leading minds to discuss the big issues facing businesses today,” the magazine says on its Twitter bio.
The magazine’s Twitter feed says it launched on March 21, though there is no mention of the magazine on Google’s blog, Twitter feed, Facebook page or newsroom.
In a note on the magazine’s website, the managing director of Google’s U.K. & Ireland Operations, Matt Brittin, said, “Think Quarterly is a breathing space in a busy world. It’s a place to take time out and consider what’s happening and why it matters.”