Story first appeared on USA Today -
Data storage shift helps companies' bottom line
Cloud computing is exploding and growing faster than a swirling funnel crossing the Oklahoma plains. The next generation of computing lowers information technology costs while increasing corporate profits at the same time. And what's not to like about that?
That one-two punch was revealed in a study obtained by USA TODAY conducted by England's Manchester Business School. The study, which was commissioned by San Antonio-based hosting company Rackspace, is expected to be released Wednesday.
The Manchester study indicates that cloud computing allows U.S. businesses to slash information technology costs by about 26%. What's more, 62% of those same American companies say that deploying in the cloud improved their bottom lines.
"The results are finally showing what we've known all along," says Rackspace Chief Technology Officer John Engates. "It's not just about moving workloads from your data center to our data center."
The rise of cloud computing has much bigger ramifications. It's a tectonic shift in how we work, live and play. iTunes is in the cloud. Ford's cars are connected to the cloud. Google's Gmail is based in the cloud. But those are largely consumer examples; now corporate computing is also shifting to the cloud.
"The move to the cloud can't happen fast enough for some companies," says Engates, who has been on the ground floor of the cloud computing movement.
Cloud computing has myriad definitions, but in the most general sense it means devices linked to data centers located just about anywhere over a combination of wireless and wired networks. There are "private clouds," where companies own and control the data centers, which are usually centrally located in lower-cost geographies. And then there are "public clouds," in which companies use computing power delivered from servers they don't own, which are usually shared with other corporate customers.
Big companies tend to use a combination of private and public clouds, reserving their high-security functions and digital record keeping for the data centers they control. But the growing acceptance of public clouds foreshadows a trend in which computing power will be delivered similarly to the way electricity is distributed by utility companies. In fact, tech geeks refer to the long-term public cloud concept as "utility computing."
We are a long way from when most companies no longer own servers, or operate so-called on-premise data centers, and rely solely on public clouds. There are a number of reasons, including security concerns, control and reliability. But the Manchester survey suggests that enterprise computer customers are embracing the shift enthusiastically.
In addition to the cost-efficiency of cloud computing, the study found that 68% of U.S. firms are plowing the cash they saved back into their businesses. They are using the cost savings to improve and expand product lines, services and other offerings. More than 60% of the companies surveyed say they are using the money to hire new employees, give raises and offer bonuses. Employment at the American companies surveyed increased 28%.
While existing companies are transitioning to cloud computing at their own pace, start-ups unsurprisingly are totally embracing the change -- especially software and social-media concerns and online retail outfits.
More than half of the start-ups surveyed said they wouldn't have been able to afford on-premise data centers at the time of their launch.
Of course, it is self-serving for a cloud-service provider to hire a study that supports its case, but the numbers are the numbers, and Manchester interviewed some 1,300 companies in both the U.S. and the United Kingdom.
Intel's general manager of cloud computing, Jason Waxman, isn't surprised by the findings. Server, storage and networking sales have been booming at the chip giant in recent years. Intel pegs the compounded growth rate for servers at about 25% to 30% a year based largely on expansion of private and public clouds.
"The more companies can save on computer infrastructure, the more they can spend on infrastructure," Waxman says. "All of these new opportunities represent a huge build-out."
Waxman thinks that public cloud providers, including Rackspace, Seattle-based Amazon.com (yes, that Amazon) and San Francisco-based GoGrid, could grow as much as 70% a year.
Gartner, the industry research consultant, predicts that the total public cloud market could swell to more than $206 billion in 2016, roughly double what it is now.
Says Intel's Waxman, "It's an astronomical opportunity."
Showing posts with label data storage. Show all posts
Showing posts with label data storage. Show all posts
Tuesday, February 26, 2013
Friday, September 3, 2010
H-P Outguns Dell in Takeover Duel
The Wall Street Journal
It wasn't the biggest bidding war the tech world has ever seen. But when it ended Thursday morning—after 10 days of all-night strategy sessions, hardball phone negotiations and power naps on couches—the battle between titans Hewlett-Packard Co. and Dell Inc. for a humble maker of data-storage systems certainly qualified as one of the wildest ever.
Winner H-P and runner-up Dell both offered fat premiums for 3PAR Inc. The company traded for less than $10 a share before Dell announced its intent to acquire it on Aug 16. H-P's winning $2.1 billion offer, or $33 a share, far surpasses Dell's original $1.1 billion bid. H-P and Dell both say they wanted 3PAR, a player in the fast-growing field of "cloud computing," to help drive their growth.
The frenzy sent 3PAR stock soaring into the stratosphere, more than tripling in just over two weeks. The bids went "into a land of dreamed-up future payoff," said Rob Cihra, an analyst at Caris & Co. In the end, he said, the bidding crossed the line into "a battle for pride more than 3PAR."
The duel was far from settled as recently as late Wednesday, said people familiar with the matter, as Dell vacillated between dropping out and pushing forward with yet another proposal for 3PAR.
Dell's bankers and lawyers had already worked around the clock for more than 48 hours on several proposals that they sent to 3PAR's advisers, snatching only a few hours of sleep in between. Late Wednesday, Dell considered withdrawing from the bidding entirely. Then it changed course, and launched a $32-a-share offer that came with a new set of conditions attached.
Even as Dell agonized, H-P's executives and advisers monitored Dell's moves through "the grapevine," said a person familiar with the matter. After hearing Dell might have a revised proposal, H-P moved ahead with a $33-a-share offer that it had prepared several days earlier to preempt a further escalation of the bidding war. H-P had been "sitting on its hands for five days," long enough to have sketched out its responses to Dell's potential moves, this person said.
Tech titans H-P and Dell have a history of feuding. They have often poached one another's executives. H-P overtook Dell as the world's largest PC maker by units in late 2006. They have both been mining the same territory to expand into new businesses.
This latest drama unfolded simultaneously in Dell's headquarters in Round Rock, Tex.; in H-P's Palo Alto, Calif., headquarters; 3PAR's offices in Fremont, Calif.; and in white-shoe investment banking offices in San Francisco and New York. Top executives such as Dell chief executive Michael Dell were kept constantly apprised on the fast-moving talks via email and phone briefings, said people familiar with the matter.
Bidding wars and hostile takeovers were once anathema in Silicon Valley, where deal-makers preferred to negotiate friendly arrangements between companies that shared similar work cultures. But as tech darlings such as Oracle Corp., Microsoft Corp. and Cisco Systems Inc. have become sprawling, maturing behemoths, their search for new sources of growth has introduced an era of more combative deal-making.
Oracle was one of the first companies to employ the hostile tactic—buying rival PeopleSoft in 2004 after an 18-month pursuit. Last year, International Business Machines Corp. was close to a deal to buy Sun Microsystems when Oracle swooped in at the last minute with a winning $7.4 billion bid. Not long after, EMC Corp. fought rival NetApp Inc. over Data Domain, a niche storage-technology company that it had coveted. Data Domain, which initially agreed to sell itself to NetApp for $25 a share, or $1.5 billion, eventually went to EMC for $33.50 per share, or about $2.4 billion.
3PAR is particularly prized because it makes storage products that are part of the field of "cloud computing," in which businesses store information in data centers operated by specialists and access that information over the Internet. A spokesman for 3PAR declined to comment.
The lofty winning offer translates to hefty riches for 3PAR investors including venture-capital firms Mayfield Fund, Menlo Ventures and Worldview Technology Partners. In addition, mutual-fund giant Fidelity Investments is among 3PAR's biggest institutional shareholders. Among individuals, 3PAR CEO David Scott, who is a former H-P executive, will reap around $100 million for his stake in the company.
Founded in 1999, 3PAR went public in 2007 and had been in the sights of H-P and Dell for some time. While 3PAR had posted losses for its past three fiscal years, it was considered a growth company—posting $194 million in revenue its fiscal year ended in March, up from $118 million two years earlier—with a hot technology.
On July 8, H-P contacted 3PAR about a potential acquisition, and executives from the two companies—including 3PAR CEO Mr. Scott and then H-P CEO Mark Hurd—met to discuss the parameters of a possible deal the next week. 3PAR's directors quickly hired Qatalyst Partners LLP, which is run by industry insider Frank Quattrone.
That same month, Qatalyst, which declined to comment, ran a mini-auction for 3PAR. In addition to Dell, which 3PAR had held discussions with about a reselling partnership, the bankers invited two more companies to join the bidding, which people familiar with the matter identified as Oracle and NetApp.
Oracle and NetApp soon dropped out, leaving Dell and H-P. In late July, H-P and Dell put in their first offers, with Dell indicating it would pay $15 to $17 a share and H-P's offer in the "mid-teens," said people familiar with the matter.
Dell soon agreed to pay $18 per share, if 3PAR entered into two weeks of exclusive negotiations, according to regulatory filings. 3PAR's advisers asked H-P if it would raise its offer, but H-P declined on Aug. 1, partly because it wasn't convinced of the seriousness of another suitor's interest in 3PAR, people familiar with the matter said.
On Aug. 16, Dell announced its intent to acquire 3PAR for $18 a share. With 62.5 million shares outstanding, the bid was valued at about $1.1 billion. That might have been the end of the story, especially since less than two weeks earlier, H-P CEO Mr. Hurd unexpectedly announced his resignation. Typically, when a company loses its chief, all strategic business is put on hold until a replacement is named. Not so in this case.
When H-P learned it was up against Dell, said people familiar with the matter, a team of executives laid out the steps they would take to thwart their rival's every possible move. A person familiar with the matter said H-P's board had pre-approved a deal for 3PAR so that the company's executives were free to make new bids without first getting clearance from the board.
On Aug. 23, H-P offered $24 per share for 3PAR.
Dell's original agreement with 3PAR gave it the right to match any counteroffer, and it came back with a $24.30-a-share bid the following day.
But H-P was "in this to win it," a person familiar with the matter said last week. Last week, H-P twice raised its bid in $3 increments, eventually offering $30 per share, or about $2 billion. "It was just adrenaline that kept people going," said one person familiar with the matter.
At Dell, mergers and strategy chief Dave Johnson, along with Mr. Dell and other senior members of the deals team and adviser Credit Suisse, considered several options to save its deal to buy 3PAR via conference calls, text messages and emails, according to people familiar with the matter. As a Wednesday deadline for Dell's response approached, some advisers pulled all-nighters at the office, sneaking in power naps on office couches.
Late Wednesday, Dell put forth its final revised proposal for 3PAR, indicating it would pay up to $32 a share if 3PAR's board agreed to include a commercial agreement and a $92 million break-up fee. But 3PAR's board balked at the idea of a commercial agreement that would have required 3PAR to sell equipment to Dell even if H-P were to buy the company, said a person familiar with the matter.
Early Thursday, Dell decided to drop out entirely, after 3PAR's board rebuffed its revised proposal yet again, people familiar with the matter said.
But H-P had heard word that Dell had a revised proposal in the works, people familiar with the matter say. H-P executives and advisers, including J.P. Morgan Chase & Co., grew worried when no announcement came. Through the night, the H-P camp pumped their sources to figure out what Dell might be planning, even as Dell and 3PAR continued to negotiate, these people said.
Around 8:30 a.m. Thursday morning, H-P representatives called 3PAR preemptively and bumped their offer to $33 a share, people familiar with the matter said. 3PAR's Mr. Scott was walking up the stairs in a London restaurant to find a quiet corner where he could dial into a board meeting to make a final decision, when the offer came in. About nine hours later, the two sides had struck a deal.
Winner H-P and runner-up Dell both offered fat premiums for 3PAR Inc. The company traded for less than $10 a share before Dell announced its intent to acquire it on Aug 16. H-P's winning $2.1 billion offer, or $33 a share, far surpasses Dell's original $1.1 billion bid. H-P and Dell both say they wanted 3PAR, a player in the fast-growing field of "cloud computing," to help drive their growth.
The frenzy sent 3PAR stock soaring into the stratosphere, more than tripling in just over two weeks. The bids went "into a land of dreamed-up future payoff," said Rob Cihra, an analyst at Caris & Co. In the end, he said, the bidding crossed the line into "a battle for pride more than 3PAR."
The duel was far from settled as recently as late Wednesday, said people familiar with the matter, as Dell vacillated between dropping out and pushing forward with yet another proposal for 3PAR.
Dell's bankers and lawyers had already worked around the clock for more than 48 hours on several proposals that they sent to 3PAR's advisers, snatching only a few hours of sleep in between. Late Wednesday, Dell considered withdrawing from the bidding entirely. Then it changed course, and launched a $32-a-share offer that came with a new set of conditions attached.
Even as Dell agonized, H-P's executives and advisers monitored Dell's moves through "the grapevine," said a person familiar with the matter. After hearing Dell might have a revised proposal, H-P moved ahead with a $33-a-share offer that it had prepared several days earlier to preempt a further escalation of the bidding war. H-P had been "sitting on its hands for five days," long enough to have sketched out its responses to Dell's potential moves, this person said.
Tech titans H-P and Dell have a history of feuding. They have often poached one another's executives. H-P overtook Dell as the world's largest PC maker by units in late 2006. They have both been mining the same territory to expand into new businesses.
This latest drama unfolded simultaneously in Dell's headquarters in Round Rock, Tex.; in H-P's Palo Alto, Calif., headquarters; 3PAR's offices in Fremont, Calif.; and in white-shoe investment banking offices in San Francisco and New York. Top executives such as Dell chief executive Michael Dell were kept constantly apprised on the fast-moving talks via email and phone briefings, said people familiar with the matter.
Bidding wars and hostile takeovers were once anathema in Silicon Valley, where deal-makers preferred to negotiate friendly arrangements between companies that shared similar work cultures. But as tech darlings such as Oracle Corp., Microsoft Corp. and Cisco Systems Inc. have become sprawling, maturing behemoths, their search for new sources of growth has introduced an era of more combative deal-making.
Oracle was one of the first companies to employ the hostile tactic—buying rival PeopleSoft in 2004 after an 18-month pursuit. Last year, International Business Machines Corp. was close to a deal to buy Sun Microsystems when Oracle swooped in at the last minute with a winning $7.4 billion bid. Not long after, EMC Corp. fought rival NetApp Inc. over Data Domain, a niche storage-technology company that it had coveted. Data Domain, which initially agreed to sell itself to NetApp for $25 a share, or $1.5 billion, eventually went to EMC for $33.50 per share, or about $2.4 billion.
3PAR is particularly prized because it makes storage products that are part of the field of "cloud computing," in which businesses store information in data centers operated by specialists and access that information over the Internet. A spokesman for 3PAR declined to comment.
The lofty winning offer translates to hefty riches for 3PAR investors including venture-capital firms Mayfield Fund, Menlo Ventures and Worldview Technology Partners. In addition, mutual-fund giant Fidelity Investments is among 3PAR's biggest institutional shareholders. Among individuals, 3PAR CEO David Scott, who is a former H-P executive, will reap around $100 million for his stake in the company.
Founded in 1999, 3PAR went public in 2007 and had been in the sights of H-P and Dell for some time. While 3PAR had posted losses for its past three fiscal years, it was considered a growth company—posting $194 million in revenue its fiscal year ended in March, up from $118 million two years earlier—with a hot technology.
On July 8, H-P contacted 3PAR about a potential acquisition, and executives from the two companies—including 3PAR CEO Mr. Scott and then H-P CEO Mark Hurd—met to discuss the parameters of a possible deal the next week. 3PAR's directors quickly hired Qatalyst Partners LLP, which is run by industry insider Frank Quattrone.
That same month, Qatalyst, which declined to comment, ran a mini-auction for 3PAR. In addition to Dell, which 3PAR had held discussions with about a reselling partnership, the bankers invited two more companies to join the bidding, which people familiar with the matter identified as Oracle and NetApp.
Oracle and NetApp soon dropped out, leaving Dell and H-P. In late July, H-P and Dell put in their first offers, with Dell indicating it would pay $15 to $17 a share and H-P's offer in the "mid-teens," said people familiar with the matter.
Dell soon agreed to pay $18 per share, if 3PAR entered into two weeks of exclusive negotiations, according to regulatory filings. 3PAR's advisers asked H-P if it would raise its offer, but H-P declined on Aug. 1, partly because it wasn't convinced of the seriousness of another suitor's interest in 3PAR, people familiar with the matter said.
On Aug. 16, Dell announced its intent to acquire 3PAR for $18 a share. With 62.5 million shares outstanding, the bid was valued at about $1.1 billion. That might have been the end of the story, especially since less than two weeks earlier, H-P CEO Mr. Hurd unexpectedly announced his resignation. Typically, when a company loses its chief, all strategic business is put on hold until a replacement is named. Not so in this case.
When H-P learned it was up against Dell, said people familiar with the matter, a team of executives laid out the steps they would take to thwart their rival's every possible move. A person familiar with the matter said H-P's board had pre-approved a deal for 3PAR so that the company's executives were free to make new bids without first getting clearance from the board.
On Aug. 23, H-P offered $24 per share for 3PAR.
Dell's original agreement with 3PAR gave it the right to match any counteroffer, and it came back with a $24.30-a-share bid the following day.
But H-P was "in this to win it," a person familiar with the matter said last week. Last week, H-P twice raised its bid in $3 increments, eventually offering $30 per share, or about $2 billion. "It was just adrenaline that kept people going," said one person familiar with the matter.
At Dell, mergers and strategy chief Dave Johnson, along with Mr. Dell and other senior members of the deals team and adviser Credit Suisse, considered several options to save its deal to buy 3PAR via conference calls, text messages and emails, according to people familiar with the matter. As a Wednesday deadline for Dell's response approached, some advisers pulled all-nighters at the office, sneaking in power naps on office couches.
Late Wednesday, Dell put forth its final revised proposal for 3PAR, indicating it would pay up to $32 a share if 3PAR's board agreed to include a commercial agreement and a $92 million break-up fee. But 3PAR's board balked at the idea of a commercial agreement that would have required 3PAR to sell equipment to Dell even if H-P were to buy the company, said a person familiar with the matter.
Early Thursday, Dell decided to drop out entirely, after 3PAR's board rebuffed its revised proposal yet again, people familiar with the matter said.
But H-P had heard word that Dell had a revised proposal in the works, people familiar with the matter say. H-P executives and advisers, including J.P. Morgan Chase & Co., grew worried when no announcement came. Through the night, the H-P camp pumped their sources to figure out what Dell might be planning, even as Dell and 3PAR continued to negotiate, these people said.
Around 8:30 a.m. Thursday morning, H-P representatives called 3PAR preemptively and bumped their offer to $33 a share, people familiar with the matter said. 3PAR's Mr. Scott was walking up the stairs in a London restaurant to find a quiet corner where he could dial into a board meeting to make a final decision, when the offer came in. About nine hours later, the two sides had struck a deal.
Thursday, May 6, 2010
Seagate Introduces Flexible FreeAgent GoFlex Storage System
PC World
Seagate has refreshed its line of portable hard drives and introduced its FreeAgent GoFlex system (available in both portable and desktop variants), which encompasses drives, connection modules, and accessories. The portable drives come in assorted capacities and colors ($100 for 320GB in silver or black; $130 for 500GB in silver, black, red, or blue; $170 for 750GB in silver or black; and $190 for 1TB in silver or black). The unique hook to these models is that Seagate has reengineered the devices to separate the drive from the bridge board that translates the drive's native SATA to another interface connection. As a result, Seagate can offer a variety of cable modules for use with a single drive--giving you plenty of connection flexibility.
This approach goes counter to the trend of putting multiple interfaces (for example, any combination of USB 2.0, USB 3.0, FireWire 400, FireWire 800, and eSATA) on a drive. One possible benefit of the new Seagate design is that you might not run into any power issues if you use the drive with an ultraportable laptop (sometimes, multiinterface drives require more power than the USB ports on such notebooks output). You could also save some money, in theory: Seagate says the new design lowers the drive's costs, as you'll pay only for the connectors you need. However, if you do need extra connectors, your costs will add up quickly--the connectors range from $30 to $80. And if you don't keep an extra around, you'll be out of luck if you take the drive somewhere and realize only later that you need a particular connection.
Seagate offers six cable kits for the portable drive system. The USB 2.0 GoFlex Cable sells for $20, the USB 3.0 cable is $30, the eSATA cable costs $20, the FireWire 800 cable is $40, and the Auto Backup cable (which turns the drive into an automatic-backup unit much like the company's now-retired Replica drive, using disk-imaging software powered by Rebit) sells for $30. If you travel a lot and you like to have a few spare cables handy, that will add up fast. It also means you have to use a different module unit for different connectors.
FreeAgent GoFlex Pro 500GBAnother potential issue, though, lies with the module concept itself. The drive has not one, but two connections to pass through (the bridge module's connector, and then the connection from the bridge module to your PC). Having two potential points of failure concerned me at first, but Seagate assured me that the connector between the drive and the module has been well tested and can withstand use. That same connector, however, was an issue on the late-beta unit I tried. I could get the module to snap securely into place, but taking the module off to swap connectors was a serious chore that made my hands hurt in the end. Seagate told me that the task shouldn't be that hard; I'll update this article with my findings when I receive a final unit. But such difficulties will be a big issue if you have to pop the module off to use it with the dock that comes with the Pro drive, for example.
Another part of Seagate's reasoning for breaking the connection module out of the drive was to make the drive more directly competitive in size with Western Digital's latest, compact My Passport series (which uses a micro-USB connector). In this respect, though, Seagate doesn't quite succeed. The FreeAgent GoFlex comes close in basic dimensions--its official listed dimensions are 4.14 by 3.23 by 0.55 inches, as compared with the Western Digital drive's 4.3 by 3.2 by 0.6 inches--but for the FreeAgent drive, those measurements don't include the cable module. In the end, the new Seagate drive design is bigger than the competition--and even worse, the cable coming out of the module is fairly stiff, which makes popping the drive into a carrying case difficult.
We tested the GoFlex Pro version (which comes with a dock and packs a 7200-rpm drive inside, instead of the standard 5400-rpm drive). This model costs $140 for 500GB and $190 for 750GB. It performed well--but surprisingly, even though the drive ran at 7200 rpm, it performed comparably to the older 5400-rpm FreeAgent Go we'd tested previously.
In our tests, we tried the 500GB drive using both the USB 2.0 and USB 3.0 modules. Over USB 2.0, the drive took 145 seconds to read a 3.68GB folder of files (same as the previous FreeAgent Go), 182 seconds to write that same folder of files (compared with 197 seconds on the older drive), 134 seconds to read a large 3.68GB file (1 second less than its predecessor), and 155 seconds to write that large file (the same as before). Over USB 3.0, the drive performed comparably to the Seagate BlackArmor PS 110, the company's first USB 3.0 drive.
Unless you require the modular flexibility this drive provides, competing models may make a more compelling choice--especially if you need to use the drive in a dock, and then remove it from the dock to take it along with you.
This approach goes counter to the trend of putting multiple interfaces (for example, any combination of USB 2.0, USB 3.0, FireWire 400, FireWire 800, and eSATA) on a drive. One possible benefit of the new Seagate design is that you might not run into any power issues if you use the drive with an ultraportable laptop (sometimes, multiinterface drives require more power than the USB ports on such notebooks output). You could also save some money, in theory: Seagate says the new design lowers the drive's costs, as you'll pay only for the connectors you need. However, if you do need extra connectors, your costs will add up quickly--the connectors range from $30 to $80. And if you don't keep an extra around, you'll be out of luck if you take the drive somewhere and realize only later that you need a particular connection.
Seagate offers six cable kits for the portable drive system. The USB 2.0 GoFlex Cable sells for $20, the USB 3.0 cable is $30, the eSATA cable costs $20, the FireWire 800 cable is $40, and the Auto Backup cable (which turns the drive into an automatic-backup unit much like the company's now-retired Replica drive, using disk-imaging software powered by Rebit) sells for $30. If you travel a lot and you like to have a few spare cables handy, that will add up fast. It also means you have to use a different module unit for different connectors.
FreeAgent GoFlex Pro 500GBAnother potential issue, though, lies with the module concept itself. The drive has not one, but two connections to pass through (the bridge module's connector, and then the connection from the bridge module to your PC). Having two potential points of failure concerned me at first, but Seagate assured me that the connector between the drive and the module has been well tested and can withstand use. That same connector, however, was an issue on the late-beta unit I tried. I could get the module to snap securely into place, but taking the module off to swap connectors was a serious chore that made my hands hurt in the end. Seagate told me that the task shouldn't be that hard; I'll update this article with my findings when I receive a final unit. But such difficulties will be a big issue if you have to pop the module off to use it with the dock that comes with the Pro drive, for example.
Another part of Seagate's reasoning for breaking the connection module out of the drive was to make the drive more directly competitive in size with Western Digital's latest, compact My Passport series (which uses a micro-USB connector). In this respect, though, Seagate doesn't quite succeed. The FreeAgent GoFlex comes close in basic dimensions--its official listed dimensions are 4.14 by 3.23 by 0.55 inches, as compared with the Western Digital drive's 4.3 by 3.2 by 0.6 inches--but for the FreeAgent drive, those measurements don't include the cable module. In the end, the new Seagate drive design is bigger than the competition--and even worse, the cable coming out of the module is fairly stiff, which makes popping the drive into a carrying case difficult.
We tested the GoFlex Pro version (which comes with a dock and packs a 7200-rpm drive inside, instead of the standard 5400-rpm drive). This model costs $140 for 500GB and $190 for 750GB. It performed well--but surprisingly, even though the drive ran at 7200 rpm, it performed comparably to the older 5400-rpm FreeAgent Go we'd tested previously.
In our tests, we tried the 500GB drive using both the USB 2.0 and USB 3.0 modules. Over USB 2.0, the drive took 145 seconds to read a 3.68GB folder of files (same as the previous FreeAgent Go), 182 seconds to write that same folder of files (compared with 197 seconds on the older drive), 134 seconds to read a large 3.68GB file (1 second less than its predecessor), and 155 seconds to write that large file (the same as before). Over USB 3.0, the drive performed comparably to the Seagate BlackArmor PS 110, the company's first USB 3.0 drive.
Unless you require the modular flexibility this drive provides, competing models may make a more compelling choice--especially if you need to use the drive in a dock, and then remove it from the dock to take it along with you.
Monday, June 15, 2009
Ice Ice Baby
Ice Energy to cool data centers
Story from C-net
Ice Energy said on Monday it has partnered with data center cooling company Data Aire to make a hybrid air conditioner that uses ice for efficient cooling.
The idea behind Ice Energy's Ice Bear air conditioner is to make ice in a storage tank during off-peak times to take advantage of lower rates. The machine uses the stored ice to make cool air during times of higher energy demand--and higher rates.
Ice Energy and Data Aire said they have tested a way to link the Ice Bear to Data Aire's cooling systems. The companies said that the combination is more efficient at cooling data centers than stand-alone systems and will cut peak electricity usage.
Ice Energy originally expected to make these refrigerator-size units for consumers, but has since focused on selling to commercial customers. California utilities PG&E, Southern California Edison, and Anaheim Public Utilities offer incentives for installation of the Ice Bear.
Wednesday, May 20, 2009
Unity Semiconductor Corp. Unveils Latest Evolution In Memory Chips
Story from the Wall Street Journal
After seven years of labor in secrecy, a Silicon Valley start-up is disclosing one of the most radical efforts yet to replace today's memory chips.
Unity Semiconductor Corp. hopes to provide an alternative to chips known as NAND flash memory, which are a mainstay of products such as digital cameras and Apple Inc.'s iPod music player. The closely held company says its technology can store four times the amount of data as NAND chips of the same size, and record data five to ten times faster.
Unity said it has working prototypes but doesn't expect to offer chips commercially for two years. Besides completing a finished product, analysts say, the start-up must find a big manufacturing partner.
"For anybody to come out with a new memory it must be adopted by one of the major NAND makers," said Alan Niebel, an analyst at Web-Feet Research who has been briefed on Unity's plans.
Unity Chief Executive Darrell Rinerson, formerly an executive at memory chip specialist Micron Technology Inc., said Unity hopes to forge a partnership with an existing memory chip maker, but didn't disclose a candidate.
NAND chips, which retain data even when electrical power is switched off, have become an essential commodity in the tech sector. But price competition has whipsawed suppliers. The research firm iSuppli estimates world-wide sales of NAND chips fell 15% in 2008 to $11.8 billion.
NAND also faces an uncertain future. As companies keep shrinking circuit dimensions, many experts believe that at some point it may become impossible to boost the storage capacity of the technology any further.
--------------------
Prominent candidates to succeed today's memory-chip technology
* PCM: Phase-change memory records data by causing material to change from a crystalline state, or phase, to a disordered state.
* MRAM: Magnetoresistive random-access memory uses magnetic charges rather than electric charges to store data.
* FRAM: Ferroelectric random-access memory is another technology for exploiting magnetic rather than electric charges.
* RRAM: Resistive random-access memory stores data as a change in the electrical resistance of materials under varying circumstances.
* NRAM: A technology based on carbon nanotubes, materials that can be fabricated in ultrasmall dimensions.
Source: WSJ Research
--------------------
Unity's approach is particularly unusual, dispensing with basic concepts such as the use of electrons to store data. The company bases its technology on ions -- charged particles formed by the addition or loss of electrons -- and the way they move through certain materials, Mr. Rinerson said.
Memory chips, which are fabricated on silicon wafers, typically store data using transistors in cells that are laid out in a two-dimensional pattern. Unity says its technology, which uses no transistors, makes it possible to stack four cells on top of each other and pack more data in less space.
In another unorthodox tactic, Unity plans to buy partially finished wafers from services called foundries. Mr. Rinerson said costly production tools will only be needed to add final layers on the wafers. As a result, a new factory to make Unity's chips might cost around $1 billion, instead of $4 billion, he said.
Unity plans to jointly finance a factory with the aid of the unnamed manufacturing partner; both companies would sell a portion of the output under their own brands. With the aid of about 60 patents issued so far, Mr. Rinerson said Unity's intent is "to keep others out, to not allow this technology to be commoditized."
Unity, which has raised close to $75 million in venture capital so far, expects by mid-2011 to offer a commercial chip that stores 64 gigabits -- about twice the capacity of the most advanced NAND chips on the data centers market now. The will help in applications where chips have played a limited role to date, Mr. Rinerson predicted.
But the company faces many hurdles, including proving it can churn out chips at competitive prices. Jim Handy, an analyst at the market-research firm Objective Analysis, said other NAND alternatives have been proposed before that never met that challenge. "The memory game is always about cost," he said.
Subscribe to:
Posts (Atom)
