Not quite Heinz: Intel's Avoton server chip comes in 13 varieties

IDG News Service - Intel is taking another run at the market for low-power, high-density servers with its new "Avoton" chip, which was launched Wednesday and will do battle with an expected upcoming wave of ARM-based processors.

About 10 vendors showed Avoton systems at an Intel event in San Francisco Wednesday. Some of them will ship later this year. They include a mix of compute, network and "cold storage" servers from Hewlett-Packard, Dell, Quanta and others.

Avoton refers to a family of chips, called the Atom C2000 line, that's built on Intel's new Silvermont core and manufactured on a 22-nanometer process. Intel says Avoton will provide seven times the performance and six times the performance per watt of its predecessor, Centerton, which came out nine months ago.

There will be 13 versions of Avoton tuned for different applications, including compute, storage, network security and wireless functions. They'll vary in the number of cores, the amount of memory they address, whether they need cooling fans, and other characteristics.

It's a big change from the past, when Intel offered basically the same server chip but at different clock speeds and with different cache sizes. And it reflects increased competition from ARM-based processors, which are being developed by several vendors for different markets.

"We're clearly demonstrating Intel's move from a general-purpose compute provider to selling optimized products that address specific workloads," said Diane Bryant, general manager of Intel's Datacenter and Connected Systems Group.

That need has become greater as vendors such as Calxeda, Marvell and Applied Micro ready ARM-based server chips. Many of the server vendors at Intel's event Wednesday said they'll also offer ARM-based systems in the coming year.

"I wouldn't call [Avoton] a game changer, but it's finally evidence that Intel takes the small-core business seriously," said industry analyst Patrick Moorhead, of Moor Insights & Strategy.

Historically, he said, Intel treated Atom as a poorer cousin to its more expensive Xeon chips, but it now seems to realize that Atom could be essential to its future.

HP said it will offer Avoton in a new version of its Moonshot server that will ship later this year. It will be targeted at front-end Web applications, dedicated hosting services and certain big data workloads, said Gerald Kleyn, director of platform engineering for HP's hyperscale servers.

"The first Moonshot system was really for static Web pages, but this gets into dynamic pages. Clearly with the higher performance you can do more work to deliver content," he said.

The server includes 45 tiny processor boards, or what HP refers to as cartridges, in a chassis that's 4.3 rack units, or about 7.5 inches (19 cm), high. That's the same number of cartridges as the existing Moonshot. But Avoton increases the core count over Centerton from two to eight, and it has two memory controllers, which improves memory bandwidth and allowed HP to attach 32GB of memory with each cartridge, Kleyn said.

Reprinted with permission from IDG.net. Story copyright 2012 International Data Group. All rights reserved.

View the original article here

Not quite Heinz: Intel's Avoton server chip comes in 13 varieties

IDG News Service - Intel is taking another run at the market for low-power, high-density servers with its new "Avoton" chip, which was launched Wednesday and will do battle with an expected upcoming wave of ARM-based processors.

About 10 vendors showed Avoton systems at an Intel event in San Francisco Wednesday. Some of them will ship later this year. They include a mix of compute, network and "cold storage" servers from Hewlett-Packard, Dell, Quanta and others.

Avoton refers to a family of chips, called the Atom C2000 line, that's built on Intel's new Silvermont core and manufactured on a 22-nanometer process. Intel says Avoton will provide seven times the performance and six times the performance per watt of its predecessor, Centerton, which came out nine months ago.

There will be 13 versions of Avoton tuned for different applications, including compute, storage, network security and wireless functions. They'll vary in the number of cores, the amount of memory they address, whether they need cooling fans, and other characteristics.

It's a big change from the past, when Intel offered basically the same server chip but at different clock speeds and with different cache sizes. And it reflects increased competition from ARM-based processors, which are being developed by several vendors for different markets.

"We're clearly demonstrating Intel's move from a general-purpose compute provider to selling optimized products that address specific workloads," said Diane Bryant, general manager of Intel's Datacenter and Connected Systems Group.

That need has become greater as vendors such as Calxeda, Marvell and Applied Micro ready ARM-based server chips. Many of the server vendors at Intel's event Wednesday said they'll also offer ARM-based systems in the coming year.

"I wouldn't call [Avoton] a game changer, but it's finally evidence that Intel takes the small-core business seriously," said industry analyst Patrick Moorhead, of Moor Insights & Strategy.

Historically, he said, Intel treated Atom as a poorer cousin to its more expensive Xeon chips, but it now seems to realize that Atom could be essential to its future.

HP said it will offer Avoton in a new version of its Moonshot server that will ship later this year. It will be targeted at front-end Web applications, dedicated hosting services and certain big data workloads, said Gerald Kleyn, director of platform engineering for HP's hyperscale servers.

"The first Moonshot system was really for static Web pages, but this gets into dynamic pages. Clearly with the higher performance you can do more work to deliver content," he said.

The server includes 45 tiny processor boards, or what HP refers to as cartridges, in a chassis that's 4.3 rack units, or about 7.5 inches (19 cm), high. That's the same number of cartridges as the existing Moonshot. But Avoton increases the core count over Centerton from two to eight, and it has two memory controllers, which improves memory bandwidth and allowed HP to attach 32GB of memory with each cartridge, Kleyn said.

Reprinted with permission from IDG.net. Story copyright 2012 International Data Group. All rights reserved.

View the original article here

Not quite Heinz: Intel's Avoton server chip comes in 13 varieties

IDG News Service - Intel is taking another run at the market for low-power, high-density servers with its new "Avoton" chip, which was launched Wednesday and will do battle with an expected upcoming wave of ARM-based processors.

About 10 vendors showed Avoton systems at an Intel event in San Francisco Wednesday. Some of them will ship later this year. They include a mix of compute, network and "cold storage" servers from Hewlett-Packard, Dell, Quanta and others.

Avoton refers to a family of chips, called the Atom C2000 line, that's built on Intel's new Silvermont core and manufactured on a 22-nanometer process. Intel says Avoton will provide seven times the performance and six times the performance per watt of its predecessor, Centerton, which came out nine months ago.

There will be 13 versions of Avoton tuned for different applications, including compute, storage, network security and wireless functions. They'll vary in the number of cores, the amount of memory they address, whether they need cooling fans, and other characteristics.

It's a big change from the past, when Intel offered basically the same server chip but at different clock speeds and with different cache sizes. And it reflects increased competition from ARM-based processors, which are being developed by several vendors for different markets.

"We're clearly demonstrating Intel's move from a general-purpose compute provider to selling optimized products that address specific workloads," said Diane Bryant, general manager of Intel's Datacenter and Connected Systems Group.

That need has become greater as vendors such as Calxeda, Marvell and Applied Micro ready ARM-based server chips. Many of the server vendors at Intel's event Wednesday said they'll also offer ARM-based systems in the coming year.

"I wouldn't call [Avoton] a game changer, but it's finally evidence that Intel takes the small-core business seriously," said industry analyst Patrick Moorhead, of Moor Insights & Strategy.

Historically, he said, Intel treated Atom as a poorer cousin to its more expensive Xeon chips, but it now seems to realize that Atom could be essential to its future.

HP said it will offer Avoton in a new version of its Moonshot server that will ship later this year. It will be targeted at front-end Web applications, dedicated hosting services and certain big data workloads, said Gerald Kleyn, director of platform engineering for HP's hyperscale servers.

"The first Moonshot system was really for static Web pages, but this gets into dynamic pages. Clearly with the higher performance you can do more work to deliver content," he said.

The server includes 45 tiny processor boards, or what HP refers to as cartridges, in a chassis that's 4.3 rack units, or about 7.5 inches (19 cm), high. That's the same number of cartridges as the existing Moonshot. But Avoton increases the core count over Centerton from two to eight, and it has two memory controllers, which improves memory bandwidth and allowed HP to attach 32GB of memory with each cartridge, Kleyn said.

Reprinted with permission from IDG.net. Story copyright 2012 International Data Group. All rights reserved.

View the original article here

AMD to ship 64-bit ARM chips to server makers next year

AMD will start shipping its first ARM server chips to manufacturers for testing in the first quarter of 2014, a company executive said on Tuesday.

"When our ARM chips start sampling, [the servers] will be [among] the first few 64-bit machines out there," said Lisa Su, senior vice president and general manager of Global Business Units at AMD.

[ Keep up on the day's tech news headlines with InfoWorld's Today's Headlines: Wrap Up newsletter. ]

AMD announced in June its first 64-bit ARM chips, code-named Seattle, which will have up to 16 CPU cores. The chips are expected to be available in servers during the second half of next year.

ARM processors are used in most smartphones and tablets, and they are being tested in servers as an alternative to the widely used x86 chips. Some believe that servers with ARM processors will be more power-efficient at processing Web applications, which would help cut electric bills. AMD says its 64-bit ARM server chips will be up to four times faster and more power-efficient than its latest quad-core x86 Opteron X-series chips, which draw up to 11 watts of power.

AMD will sell the 64-bit ARM processors alongside x86 server processors, but the company expects ARM to outsell x86 in the long run. In 2011, ARM announced the first 64-bit ARMv8 CPU architecture, and AMD will be one of the first to ship server chips based on that technology. Most of the current ARM chips are just 32-bit, while x86 server chips from Intel and AMD have a head start with 64-bit support.

AMD is being patient with ARM. Adoption of the new architecture will take time as it will require a change in the way servers are built and software is written, Su said.

"Our thought process is it's a multiyear effort for the market to grow. But it's important to get the software development on it," Su said. "We definitely view it as a long-term investment."

There is a lot of interest in ARM chips, especially for dense servers, Su said. AMD last year bought SeaMicro, which offers dense servers packed with low-power x86 chips that can scale up performance depending on the workload.

AMD is also looking to expand into new markets with its custom-chip business, which builds device-specific hardware for customers. AMD has said the custom-chip business could deliver as much as 20 percent of the company's revenue by the fourth quarter, which closes in December, and more than half the revenue in the coming years.

The custom-chip business is already showing results, with Sony and Microsoft using AMD's CPU and GPU technology for their upcoming gaming consoles. The company is also targeting the living room through set-top boxes and other devices that process high-definition video and deliver content from the cloud.

"As we look forward, there are a number of areas where the same model can work. They are in the area of specialty tablets, living room as we look through how the living room evolves, and specialty server applications," Su said.

But AMD isn't losing sight of the PC market, which continues to be its main revenue driver. The PC market has collapsed as tablet and smartphone adoption grows, but the lines are blurring between hybrids, tablets and laptops, Su said.

The company's latest Temash chips and Kabini chips are making their way to Windows tablets and laptops starting at US$399.

"We continue to believe Windows is very important. It's very capable. Windows 8 is something we are committed to, but we also believe Android is important as well," Su said.

AMD has virtually no market share in tablets, and Su acknowledged that the tablet chip is a work in progress. She also pointed out there's potential to grow in the server and custom-chip markets. After years of losses, the return to profitability in the second quarter was a key step ahead, and the company now has a sustainable business model.

"It's very important for us to continue to execute," Su said.

Agam Shah covers PCs, tablets, servers, chips and semiconductors for IDG News Service. Follow Agam on Twitter at @agamsh. Agam's e-mail address is agam_shah@idg.com

View the original article here

AMD to ship 64-bit ARM chips to server makers in 2014

IDG News Service - Advanced Micro Devices will start shipping its first ARM server chips to manufacturers for testing in the first quarter of 2014, a company executive said on Tuesday.

"When our ARM chips start sampling, [the servers] will be [among] the first few 64-bit machines out there," said Lisa Su, senior vice president and general manager of Global Business Units at AMD.

AMD announced in June its first 64-bit ARM chips, code-named Seattle, which will have up to 16 CPU cores. The chips are expected to be available in servers during the second half of next year.

ARM processors are used in most smartphones and tablets, and are being tested in servers as an alternative to the widely used x86 chips. Some believe that servers with ARM processors will be more power-efficient at processing Web applications, which would help cut electric bills. AMD says its 64-bit ARM server chips will be up to four times faster and more power-efficient than its latest quad-core x86 Opteron X-series chips, which draw up to 11 watts of power.

AMD will sell the 64-bit ARM processors alongside x86 server processors, but the company expects ARM to outsell x86 in the long run. In 2011, ARM announced the first 64-bit ARMv8 CPU architecture, and AMD will be one of the first to ship server chips based on that technology. Most of the current ARM chips are just 32-bit, while x86 server chips from Intel and AMD have a head start with 64-bit support.

AMD is being patient with ARM. Adoption of the new architecture will take time as it will require a change in the way servers are built and software is written, Su said.

"Our thought process is it's a multiyear effort for the market to grow. But it's important to get the software development on it," Su said. "We definitely view it as a long-term investment."

There is a lot of interest in ARM chips, especially for dense servers, Su said. AMD last year bought SeaMicro, which offers dense servers packed with low-power x86 chips that can scale up performance depending on the workload.

AMD is also looking to expand into new markets with its custom-chip business, which builds device-specific hardware for customers. AMD has said the custom-chip business could deliver as much as 20 percent of the company's revenue by the fourth quarter, which closes in December, and more than half the revenue in the coming years.

The custom-chip business is already showing results, with Sony and Microsoft using AMD's CPU and GPU technology for their upcoming gaming consoles. The company is also targeting the living room through set-top boxes and other devices that process high-definition video and deliver content from the cloud.

Reprinted with permission from IDG.net. Story copyright 2012 International Data Group. All rights reserved.

View the original article here

AMD to ship 64-bit ARM chips to server makers next year

Advanced Micro Devices will start shipping its first ARM server chips to manufacturers for testing in the first quarter of 2014, a company executive said on Tuesday.

“When our ARM chips start sampling, [the servers] will be [among] the first few 64-bit machines out there,” said Lisa Su, senior vice president and general manager of Global Business Units at AMD.

AMD announced in June its first 64-bit ARM chips, code-named Seattle, which will have up to 16 CPU cores. The chips are expected to be available in servers during the second half of next year.

ARM processors are used in most smartphones and tablets, and are being tested in servers as an alternative to the widely used x86 chips. Some believe that servers with ARM processors will be more power-efficient at processing Web applications, which would help cut electric bills. AMD says its 64-bit ARM server chips will be up to four times faster and more power-efficient than its latest quad-core x86 Opteron X-series chips, which draw up to 11 watts of power.

AMD will sell the 64-bit ARM processors alongside x86 server processors, but the company expects ARM to outsell x86 in the long run. In 2011, ARM announced the first 64-bit ARMv8 CPU architecture, and AMD will be one of the first to ship server chips based on that technology. Most of the current ARM chips are just 32-bit, while x86 server chips from Intel and AMD have a head start with 64-bit support.

AMD is being patient with ARM. Adoption of the new architecture will take time as it will require a change in the way servers are built and software is written, Su said.

“Our thought process is it’s a multiyear effort for the market to grow. But it’s important to get the software development on it,” Su said. “We definitely view it as a long-term investment.”

There is a lot of interest in ARM chips, especially for dense servers, Su said. AMD last year bought SeaMicro, which offers dense servers packed with low-power x86 chips that can scale up performance depending on the workload.

AMD is also looking to expand into new markets with its custom-chip business, which builds device-specific hardware for customers. AMD has said the custom-chip business could deliver as much as 20 percent of the company’s revenue by the fourth quarter, which closes in December, and more than half the revenue in the coming years.

The custom-chip business is already showing results, with Sony and Microsoft using AMD’s CPU and GPU technology for their upcoming gaming consoles. The company is also targeting the living room through set-top boxes and other devices that process high-definition video and deliver content from the cloud.

“As we look forward, there are a number of areas where the same model can work. They are in the area of specialty tablets, living room as we look through how the living room evolves, and specialty server applications,” Su said.

But AMD isn’t losing sight of the PC market, which continues to be its main revenue driver. The PC market has collapsed as tablet and smartphone adoption grows, but the lines are blurring between hybrids, tablets and laptops, Su said.

The company’s latest Temash chips and Kabini chips are making their way to Windows tablets and laptops starting at $399.

“We continue to believe Windows is very important. It’s very capable. Windows 8 is something we are committed to, but we also believe Android is important as well,” Su said.

AMD has virtually no market share in tablets, and Su acknowledged that the tablet chip is a work in progress. She also pointed out there’s potential to grow in the server and custom-chip markets. After years of losses, the return to profitability in the second quarter was a key step ahead, and the company now has a sustainable business model.

“It’s very important for us to continue to execute,” Su said.

Agam Shah is a reporter for the IDG News Service in New York. He covers hardware including PCs, servers, tablets, chips, semiconductors, consumer electronics and peripherals.
More by Agam Shah, IDG News Service

View the original article here

Imagination to battle ARM with MIPS server chips

Imagination Technologies, the UK company that designs the graphics chips for the iPhone and iPad, will release a new CPU design next year for low-power servers, the company's CEO said this week.

It's part of a broad push by Imagination to expand the use of MIPS chips in all kinds of products, including servers, tablets, smartphones and wearable computers. Imagination inherited the architecture when it bought financially struggling MIPS Technologies last year.

[ Keep up on the day's tech news headlines with InfoWorld's Today's Headlines: Wrap Up newsletter. ]

"The intent with MIPS is to cover all the markets where a CPU is relevant," said Hossein Yassaie, Imagination's CEO, at a press event in Silicon Valley on Wednesday.

It's a new and bold effort by Imagination, which is best known for the PowerVR graphics chips used in many top-tier mobile devices, including Google Glass. It hopes to expand the MIPS business in areas where the architecture is already strong, such as networking and home media systems, and also to plow new ground.

That will mean increased competition with Imagination's UK rival ARM Holdings. ARM is the dominant designer of CPUs for smartphones and tablets, and it's also targeting the server market, so the moves by Imagination could set up a grand battle between the two companies over the next few years. ARM, meanwhile, is also going for a bigger slice of Imagination's graphics business.

In servers, both companies are targeting low-power "hyperscale" servers, used by companies such as Facebook and Google to run massive online workloads. They will both have to compete with incumbent Intel, which has stepped up its own efforts to build low-power chips.

ARM and Imagination both develop chip designs that they license to other companies to manufacture. They each offer architectural licenses, which let companies design their own chips from scratch, and licenses for complete core designs, which can be turned into finished products more quickly.

Companies such as Broadcom and Cavium already build MIPS chips for specialized network equipment using an architectural license. But to promote MIPS more broadly, Imagination will offer complete core designs for servers as well.

Asked to give a timeframe, Yassaie said Imagination's upcoming Series5 "Warrior" CPUs will cover every segment of the market -- including servers -- by the end of 2014. It will take a further two to three years for chip makers to create finished server products and get them to market, Yassaie said.

"So for us to be a significant player in [servers] it will take several years. But in terms of technology, come the end of next year I would expect to have a solution for all the markets that are important for a CPU to be in," he said.

Imagination had already made clear that it would use the MIPS acquisition to build CPUs for the tablet and smartphone markets. Yassaie's comments Wednesday show the company is serious about servers as well.

View the original article here

Cisco fixes critical remote command execution vulnerability in Secure Access Control Server

Cisco Systems released security patches for Secure Access Control Server (Secure ACS) for Windows to address a critical vulnerability that could allow unauthenticated attackers to remotely execute arbitrary commands and take control of the underlying operating system.

Cisco Secure ACS is an application that allows companies to centrally manage access to network resources for various types of devices and users. According to Cisco's documentation, it enforces access control policies for VPN, wireless and other network users and it authenticates administrators, authorizes commands, and provides an audit trail.

[ InfoWorld's expert contributors show you how to secure your Web browsers in a free PDF guide. Download it today! | Learn how to protect your systems with Roger Grimes' Security Adviser blog and Security Central newsletter, both from InfoWorld. ]

Cisco Secure ACS supports two network access control protocols: Remote Access Dial In User Service (RADIUS) and Terminal Access Controller Access-Control System Plus (TACACS+).

The newly patched vulnerability is identified as CVE-2013-3466 and affects Cisco Secure ACS for Windows versions 4.0 through 4.2.1.15 when configured as a RADIUS server with Extensible Authentication Protocol-Flexible Authentication via Secure Tunneling (EAP-FAST) authentication.

"The vulnerability is due to improper parsing of user identities used for EAP-FAST authentication," Cisco said Wednesday in a security advisory. "An attacker could exploit this vulnerability by sending crafted EAP-FAST packets to an affected device."

"Successful exploitation of the vulnerability may allow an unauthenticated, remote attacker to execute arbitrary commands and take full control of the underlying operating system that hosts the Cisco Secure ACS application in the context of the System user for Cisco Secure ACS running on Microsoft Windows," the company said.

The vulnerability received the maximum severity score, 10.0, in the Common Vulnerability Scoring System (CVSS), which indicates that it is highly critical. Cisco Secure ACS for Windows version 4.2.1.15.11 was released to address the flaw.

There are no known workarounds, so upgrading to the patched version of the application is recommended.

View the original article here

Imagination to battle ARM with MIPS server chips

IDG News Service - Imagination Technologies, the UK company that designs the graphics chips for the iPhone and iPad, will release a new CPU design next year for low-power servers, the company's CEO said this week.

It's part of a broad push by Imagination to expand the use of MIPS chips in all kinds of products, including servers, tablets, smartphones and wearable computers. Imagination inherited the architecture when it bought financially struggling MIPS Technologies last year.

"The intent with MIPS is to cover all the markets where a CPU is relevant," said Hossein Yassaie, Imagination's CEO, at a press event in Silicon Valley on Wednesday.

It's a new and bold effort by Imagination, which is best known for the PowerVR graphics chips used in many top-tier mobile devices, including Google Glass. It hopes to expand the MIPS business in areas where the architecture is already strong, such as networking and home media systems, and also to plow new ground.

That will mean increased competition with Imagination's UK rival ARM Holdings. ARM is the dominant designer of CPUs for smartphones and tablets, and it's also targeting the server market, so the moves by Imagination could set up a grand battle between the two companies over the next few years. ARM, meanwhile, is also going for a bigger slice of Imagination's graphics business.

In servers, both companies are targeting low-power "hyperscale" servers, used by companies such as Facebook and Google to run massive online workloads. They will both have to compete with incumbent Intel, which has stepped up its own efforts to build low-power chips.

ARM and Imagination both develop chip designs that they license to other companies to manufacture. They each offer architectural licenses, which let companies design their own chips from scratch, and licenses for complete core designs, which can be turned into finished products more quickly.

Companies such as Broadcom and Cavium already build MIPS chips for specialized network equipment using an architectural license. But to promote MIPS more broadly, Imagination will offer complete core designs for servers as well.

Asked to give a timeframe, Yassaie said Imagination's upcoming Series5 "Warrior" CPUs will cover every segment of the market -- including servers -- by the end of 2014. It will take a further two to three years for chip makers to create finished server products and get them to market, Yassaie said.

"So for us to be a significant player in [servers] it will take several years. But in terms of technology, come the end of next year I would expect to have a solution for all the markets that are important for a CPU to be in," he said.

Reprinted with permission from IDG.net. Story copyright 2012 International Data Group. All rights reserved.

View the original article here

Imagination to battle ARM with MIPS server chips

Imagination Technologies, the U.K. company that designs the graphics chips for the iPhone and iPad, will release a new CPU design next year for low-power servers, the company’s CEO said this week.

It’s part of a broad push by Imagination to expand the use of MIPS chips in all kinds of products, including servers, tablets, smartphones. and wearable computers. Imagination inherited the architecture when it bought financially struggling MIPS Technologies last year.

“The intent with MIPS is to cover all the markets where a CPU is relevant,” said Hossein Yassaie, Imagination’s CEO (shown above), at a press event in Silicon Valley on Wednesday.

It’s a new and bold effort by Imagination, which is best known for the PowerVR graphics chips used in many top-tier mobile devices, including Google Glass. It hopes to expand the MIPS business in areas where the architecture is already strong, such as networking and home media systems, and also to plow new ground.

That will mean increased competition with Imagination’s UK rival ARM Holdings. ARM is the dominant designer of CPUs for smartphones and tablets, and it’s also targeting the server market, so the moves by Imagination could set up a grand battle between the two companies over the next few years. ARM, meanwhile, is also going for a bigger slice of Imagination’s graphics business.

In servers, both companies are targeting low-power “hyperscale” servers, used by companies such as Facebook and Google to run massive online workloads. They will both have to compete with incumbent Intel, which has stepped up its own efforts to build low-power chips.

ARM and Imagination both develop chip designs that they license to other companies to manufacture. They each offer architectural licenses, which let companies design their own chips from scratch, and licenses for complete core designs, which can be turned into finished products more quickly.

Companies such as Broadcom and Cavium already build MIPS chips for specialized network equipment using an architectural license. But to promote MIPS more broadly, Imagination will offer complete core designs for servers as well.

Asked to give a timeframe, Yassaie said Imagination’s upcoming Series5 “Warrior” CPUs will cover every segment of the market—including servers—by the end of 2014. It will take a further two to three years for chip makers to create finished server products and get them to market, Yassaie said.

“So for us to be a significant player in [servers] it will take several years. But in terms of technology, come the end of next year I would expect to have a solution for all the markets that are important for a CPU to be in,” he said.

Imagination had already made clear that it would use the MIPS acquisition to build CPUs for the tablet and smartphone markets. Yassaie’s comments Wednesday show the company is serious about servers as well.

There’s nothing about ARM that makes it intrinsically more suitable for servers than MIPS, said Paul Teich, senior analyst with Moor Insights and Strategy. Both are based on RISC (Reduced Instruction Set Computer) architectures that lend themselves to low-power designs.

The “magic” in ARM server chips from the likes of Calxeda doesn’t come from the core itself but from the surrounding technologies, Teich said. Nor should software be a hurdle for MIPS, he said, since there are already MIPS versions of Linux, and it would be relatively easy to port other programs to the architecture.

Gartner analyst Mark Hung agreed that “from a technical standpoint,” MIPS can be competitive with ARM in servers.

But ARM has a head start. Calxeda and Applied Micro have already developed ARM server chips that customers are testing.

If the new MIPS cores don’t appear until the end of next year, Imagination will be “on the tail end of being late,” Teich said.

Yassaie acknowledged there’s work to be done in both servers and smartphones. Initially, he expects to grow the MIPS business in markets where it’s already prevalent, such as networking and home media equipment. Mobile may be the hardest market for Imagination to crack, he said, presumably because of ARM’s dominance.

But the industry is ready for another CPU design company, according to Yassaie. Vendors don’t want to be beholden to a single supplier, and he’s quick to label ARM a “monopoly” for its dominance in mobile devices.

Several years ago, the PC was “the only ecosystem that existed,” he said. “But now there are several ecosystems, with some pretty big players behind them, and it’s completely untenable to expect those people to put up with such control.”

There’s also renewed confidence in MIPS now that it’s owned by a stable company, according to Yassaie. “People were worried about MIPS’s future, there was a cloud over its head, and now that’s gone away.”

View the original article here

Server revenue continued slide in Q2 on weak demand, says IDC

The server business continued to slide in the second quarter with worldwide revenue and unit sales down, IDC said Tuesday.

Revenue was down 6.2 per cent to $11.9 billion in the second consecutive quarter of year-over-year decline, as demand for servers continued to soften in most geographic regions, the research firm said. Unit shipments were also down 1.2 percent to 2 million, after also falling in the previous two quarters.

[ Get the latest practical info and news with Paul Venezia's The Deep End blog and InfoWorld's Data Center newsletter. ]

The highest fall in revenue was in midrange systems, which dipped by about 22 percent year-over-year, while volume systems had a 2.4 percent revenue decline and revenue from high-end systems dipped 9.5 percent in the quarter ended June.

The slowing demand is a combination of factors including consolidation, virtualization, and migration initiatives by mainstream small and medium businesses and enterprise customers, and dampening demand for new IT projects in difficult economic conditions. Top server vendors tried to offset weak demand for higher-margin Unix and blade servers with lower-margin rack and density optimized servers, IDC said.

IBM held the number one position in the server market with a 27.9 percent share of revenue, but its share was down from over 29 percent last year. The company's server revenue fell in the quarter by 10 percent year-over-year because of low demand for System x and Power Systems. IBM's System z mainframe running z/OS, however, had a third consecutive quarter of growth, with revenue up by 9.9 percent year-over-year to $1.2 billion. The mainframe accounted for 9.8 percent of server revenue in the quarter.

Hewlett-Packard held the number two position with a smaller market share of close to 26 percent, after a year-on-year decline in revenue of 17.5 percent as a result of poor demand for x86-based ProLiant servers and continued decline in demand for HP Integrity servers.

Dell fared better, growing its server market share to 18.8 percent, its highest ever in any quarter, from 16 percent last year, to retain the third spot. Dell's server revenue grew 10.3 percent in the quarter. Rival Oracle held the fourth position with 6 percent market share followed by Cisco Systems at number five with 4.5 percent share after a close to 43 percent growth in revenue. Cisco had a statistical tie with Fujitsu in the last quarter.

Revenue from density optimized servers, used in large hyperscale datacenters, surged 26.6 percent year-on-year in the quarter to $735 million, while units shipped increased 13.8 percent to close to 200,000 servers. Dell continues to lead in this market with a whopping 60.5 percent share.

In contrast, revenue from blade servers was down 6.2 percent year-on-year to $2.0 billion. Blades accounted for close to 17 percent of total server revenue in the quarter. HP held the top position in this market with 44.8 percent share of revenue.

Demand for x86 servers was weak in the quarter, with revenue down by 1.3 percent year-over-year to $8.7 billion as shipments were relatively flat at 1.9 million servers, according to IDC.

Linux servers benefited from cloud infrastructure deployments and now account for about 23 percent of all server revenue, up by 1.8 percentage points from the same quarter a year ago. Windows server sales were, however, down 5.1 percent, with quarterly server hardware revenue at $5.8 billion, representing 49.3 percent of overall quarterly server revenue, IDC said. Unix server revenue declined 21 percent to its lowest quarterly revenue of $1.8 billion, accounting for about 15 percent of server revenue for the quarter.

John Ribeiro covers outsourcing and general technology breaking news from India for The IDG News Service. Follow John on Twitter at @Johnribeiro. John's email address is john_ribeiro@idg.com.

View the original article here

Intel to ship new Atom server processors next week

Sorry, I could not read the content fromt this page.

View the original article here

Windows Store ad firm sees evidence of new Surface slates, Nokia tablet in its server logs

A new generation of Surface slates may be headed our way, if a new report by advertising firm AdDuplex is any indication, and Microsoft's self-made tablets may be bringing a friend along with them: A Windows RT device crafted by Nokia.

News of refreshed Surface tablets isn't exactly shocking, as Nvidia has confirmed that a next-gen Surface RT slate in the works and Microsoft has been steeply discounting the first generation Surface tablets in a bid to clear stock room shelves.

AdDuplex claims to "regularly see devices named Microsoft Corporation Surface 2 and Microsoft Corporation Surface with Windows 8.1 Pro" on its network, which provides advertising for 393 Windows Store apps. The first device looks to be a completely new model, AdDuplex claims, but the new Surface Pro device might be a first-gen Surface Pro running an RTM version of Windows 8.1.

That said, the firm claims prior preview versions of the Windows 8.1 update didn't change the reported name of the Pro, and a refreshed Surface Pro is almost a given. The original tablet suffered from poor battery life and a slightly chunky design, and those two problems could be significantly reduced if a Surface Pro with Intel's new power-sipping Haswell processors is released.

Digi-wo.comAn alleged image of Nokia's Windows RT slate, as leaked on digi-wo.com.

Repeated rumors have also had Nokia working on a Windows RT device, and AdDuplex claims that Nokia-made Windows 8/RT devices have been pinging its servers since last October.

The device was named RX-107. Then we’ve seen RX-108. And then RX-113, 114, 115, 116. Interestingly enough we haven’t seen RX-109 through RX-112. So it’s hard to say if all of these are just variations of one device progressing through development stages, or 2 device lines (107-108 and 113-116) or something else .

Monday morning, The Verge reported that Nokia's slate is dubbed "Sirius" and will resemble an oversized version of the company's popular Lumia line of Windows Phones. According to the site's sources, the Sirius tablet will be thinner and lighter than an iPad, with a 10.1-inch 1080p display, 2GB of RAM, and a cutting-edge Qualcomm Snapdragon 800 processor. Even better: Nokia is reportedly shooting for 10 hours of battery life.

If Nokia does release a Windows RT slate of its own, it'll have an uphill climb: Windows RT itself is tanking hard, and thus far, third-party Windows tablets—especially Windows RT tablets—are selling even worse than the Surface line.

Brad Chacos spends the days jamming to Spotify, digging through desktop PCs and covering everything from BYOD tablets to DIY tesla coils.
More by Brad Chacos

View the original article here

Nginx Web server goes commercial with new release

Nginx is releasing a commercial edition of its namesake open source Web server software, called Nginx Plus, which it will market as a software alternative to application delivery controllers (ADCs).

"The features that have been added are very similar to what you'd find in hardware-based ADCs," said Gus Robertson, CEO of Nginx. "We're an alternative to Apache, but I think we're also an alternative to hardware-based application delivery controllers."

Robertson said that the software will provide functionality similar to that of ADC appliances from F5 Networks or Citrix, which are used to speed delivery of high traffic websites through techniques of load balancing and caching.

Nginx has already made a name for itself as a Web server able to very efficiently handle large amounts of traffic.

Russian system administrator Igor Sysoev released the first version of Nginx in 2004 under a BSD open source license, after becoming frustrated by what he felt were the limitations of the market-leading Apache Web server software.

Sysoev specifically designed Nginx (pronounced "Engine X") for handling large volumes of traffic, up to 10,000 concurrent connections per server. It includes advanced traffic management features often found in ADCs, such as load balancing, edge caching and reverse proxy services.

According to the company, Nginx is now the most widely used Web server software among the world's 1,000 busiest sites. Web heavyweights such as Netflix, Hulu, Pinterest, AirBnB, WordPress.com, GitHub, SoundCloud, Zynga, Eventbrite and Zappos have all used Nginx to serve their sites.

The software now runs about 100 million websites overall, or about 14.55 percent of the Web, according to the latest Internet survey from Netcraft.

Sysoev founded Nginx the company in 2011, and took the role of chief technology officer. Nginx attracted US$3 million in first-round venture capital funding. Currently, it employs about 15 people. Robertson took over as CEO in April; formerly, he was the vice president of global business development for Red Hat, another open source software company.

Heretofore, Nginx, the company, has chiefly offered subscription support services for the open source software, around configuration and performance tuning, performance optimization and technical account management. Nginx Plus is the company's first commercial product.

Under a dual license, the commercial version of Nginx will offer additional features over the freely available open source version.

"The advanced functionality enhances the capabilities of Nginx in regards to load balancing, request routing, health monitoring and general control and monitoring of Nginx instances in mission critical environments," said Andrew Alexeev, Nginx co-founder and head of business development.

One of the most significant new features is the ability to make configuration changes to a running copy of Nginx, without the need to restart the software for the changes to take affect. Typically configuration changes to Web server software, such as Nginx or Apache, require the software to be stopped and then restarted, which can momentarily disrupt operations.

"Even graceful restarts can affect the operation, and adds slight overhead. [Users] want to change aspects of Nginx on the fly," Alexeev said.

With monitoring, the commercial package provides metrics of a running copy of Nginx, in the JSON (JavaScript Simple Object Notation) format. "You can import these metrics into any standard commercial or open source monitoring system, or you could use your own template through HTML or JavaScript," Alexeev said. The software package also comes with its own metrics dashboard as well.

The health check can alert an administrator when a back-end component of a Web application has failed, such as the database. It does this by comparing the size of a new Web page with the size it should be. It can also automatically take out of service those servers relying on the failing components. Nginx has also collaborated with New Relic to integrate the data generated by health checks to show up in New Relic's application performance management (APM) software.

Other features of the commercial edition include additional load balancing capabilities, and advanced media streaming for Adobe HDS and Apple HLS video formats.

Although Nginx Plus should run on any standard Linux distribution, Nginx has certified that the software runs on Amazon Linux, Red Hat, CentOS, Ubuntu, and Debian distributions.

Nginx Plus will cost $1,350 per instance per year and is available now.

Joab Jackson covers enterprise software and general technology breaking news for The IDG News Service. Follow Joab on Twitter at @Joab_Jackson. Joab's e-mail address is Joab_Jackson@idg.com

View the original article here

Rackspace hosts VMware management with new dedicated server

The introduction of Rackspace’s hosted Dedicated VMware vCenter Server will allow IT staff to control their VMware environments from a data center run by the vendor.

As enterprises move IT infrastructure out of their own data centers, vendors are offering a growing list of alternatives that in Rackspace’s world includes hybrid clouds, which combine public and private clouds, and dedicated hosting. The latest addition to the latter offering is Dedicated VMware vCenter Server, which allows IT departments to retain full control with the tools they are used to without having to bother with the underlying infrastructure.

“Many of our customers have large VMware installations in-house today, and they have made significant investments in that and don’t want to throw it away,” said Andrew Wing, senior product manager at Rackspace.

For companies that want to stick with VMware, but don’t want to expand their data centers or have data centers at all, Rackspace already offers managed virtualization based on VMware vSphere 5.1. The addition of vCenter builds on that. From a single console, administrators can control virtual servers running in-house and in Rackspace’s data centers. It is also possible to mix vCenter servers that run in an enterprise’s own data center and ones that are hosted by Rackspace.

Until now, Rackspace’s hosted VMware platform used a shared vCenter model, which only offered limited management capabilities, according to Wing.

Rackspace will at first charge a yet undecided monthly fee per hypervisor that is managed by the hosted vCenter server and in the next phase move to a model where enterprises pay for what they use.

“It is just a question of getting our billing systems set up to make that a reality,” Wing said.

Today, enterprises are using Rackspace’s managed virtualization to run things like e-commerce platforms and Web content management systems. But Rackspace is hoping to attract more central IT functions, including ERP and CRM systems, according to Wing.

View the original article here

Microsoft puts out "HTTP 2.0" server to test the future of the Web

Microsoft this week released “Katana,” a server implementation that will allow developers to begin trying out Microsoft’s version of what could become HTTP 2.0: the future of the Web.

A Microsoft blog post provides the endpoints that can be used in conjunction with an HTTP 2.0-compliant browser, or users can simply download the test code itself and try it out.

Most people rarely consider the incredible complexity of navigating to a specific site on the Web: how the Web address is looked up, translated, and the back-and-forth as data is requested and delivered.

Most of the Web still runs on on HTTP 1.0, the original spec; what’s important is that HTTP 1.0 processes one request at a time. HTTP 1.1, a slight improvement, can still end up clogging the TCP “pipe” when multiple requests are made on a single new TCP connection. HTTP 2.0 hopes to change that, so that multiple page requests can be processed via a single connection to the server, easing the load on the server and the network connecting it.

HTTP 2.0 isn’t a magic bullet, however. What consumers care about is usually just how quickly a webpage loads, however, and there are a number of technologies that, together, can improve it. These include:

Preloading or prefetching pages, as modern browsers like Google’s Chrome and Microsoft’s Internet Explorer do, Rendering them as quickly as possible, andInfrastructure improvements such as faster throughput and HTTP, the negotiation process between webpage and browser.

Enter HTTP 2.0. So far, the draft implementation of HTTP 2.0 has been based on SPDY, the open networking protocol developed by Google, which is supported by all the major browsers, save Safari. SPDY multiplexes webpage requests, so that only one connection is required.

Microsoft , meanwhile, began publishing its own approach last year, known as ”HTTP Speed + Mobility”. Basically, Microsoft believes that HTTP should be optimized not just for browsers, but for applications that access webpages and services, as well as mobile devices and browsers.

At this point, Microsoft and Google have until April 2014 to convince the Internet Engineering Task Force (IETF) to adopt either SPDY, Microsoft’s HTTP Speed + Mobility, or some combination of the two. By November, the IETF is scheduled to submit whatever compromise is hammered out to the Internet Engneering Standards Group as a formal standard.

Whichever standard is implemented should have little impact on users, as it will undoubtedly be quickly supported in various browsers, both for mobile and desktop.

But being able to take credit for architecting the next foundation of the Web would be a nice trophy for either Google or Microsoft, even if the standard should theoretically be supported by the Web at large. If nothing else, it’s another signal that HTTP 2.0 is on its way.

View the original article here