Posts Tagged ‘servers’

« Older Entries

EU eyes energy-efficient design of computer servers

February 17th, 2012
Share

Computer servers, the high-performance machines used to run programmes and software services, are facing a design revamp that will make them more energy efficient, following recent assessments by the European Commission.

Servers, together with data storage equipment and four other groups of products, have recently made the Commission’s “priority list” for the Ecodesign Directive because of their “significant” energy savings potential.

The Commission said the combined energy saving of these products would amount to 1157 TWh per year by 2030 – twice the total annual energy consumption in Sweden.

“If they know they will be regulated, it puts more pressure on [industry] to go green,” said Edouard Toulouse of the European Environmental Citizens Organisation for Standardisation (ECOS), who has been taking part in discussions on the directive.

Big players of cloud computing – like Microsoft, Google, Facebook and Amazon – could lead the industry by opening the discussion on server efficiency.

“It’s in their interest to have energy efficient computers and they are constructive and progressive about it,” Toulouse told EurActiv.

More efficient servers and data-storage equipment might mean fewer costs for companies and lower bills for consumers, the European Consumers Organisation BEUC says.

Stakeholders were to meet again this week for a discussion on the next wave of products to be regulated under the Ecodesign Directive.

The Commission needs to assess in the coming months the efforts already made on energy-efficient servers and data storage equipment and decide if there are any issues to address.

“It pushes them to start looking into this matter, and if voluntary approach is deemed satisfactory, there will be no need for regulation,” Toulouse said.

If it identifies shortcomings that cannot be resolved through voluntary agreements from the industry, then the EU executive will act to set laws within three years.

Uptake of cloud computing

Cloud computing, which refers to software and data stored remotely by companies or individuals, requires service provides to have vast server and storage capacity.

“There is still a big problem with cloud. People don’t use it enough, even if companies like Microsoft are trying to make it seem as simple as possible,” a software programmer working for an international computing company told EurActiv.

“But once they realise they can address the need to buy a special research programme and hire additional staff for a one-off situation, or without their servers crashing, for example, they will want to instead ‘rent’ this service through cloud computing”.

Industry ahead of the wave

As cloud providers prepare for the growth of services, they are also getting ready for the additional expenses they would incur – including the cost of energy to run the servers.

The market is already ahead of this development – which increases its chances to reach a voluntary agreement on energy-efficient servers with the Commission.

For example, Microsoft is leading discussions on energy efficiency and Google has created incentives to save as much energy as possible in the power management and cooling of data centres that store servers.

One of the latter’s most recent examples is the data centre set up on the Finnish coast, in Hamina, where water is pumped through an underground tunnel to cool the building.

But cooling data centres is not the top priority of these companies – energy-efficient conversion servers and methods of power conversion have the best energy savings potential.

An investment of €19,000 in energy-saving design generates returns of €50,000 per year, Google reports.

Manufacturers disagree

In the coming weeks, stakeholders will comment on possible obstacles in the inclusion of servers and data storage equipment in the Ecodesign Directive. One of the thorny issues is the initial cost required to make improvements.

Some industry voices say they are already working with the Commission on developing ecodesign requirements for some server systems.

“The Commission needs to weigh this up, along with the fact that the data centre industry is moving way faster than regulation, in evaluating whether an ecodesign measure is the best way forward,” said Kirsty Macdonald, senior energy policy manager at Intel Corporation.

BEUC’s position is that the industry has been delaying the implementation of binding EU measures and is refraining from new legislation. “The ICT industry has been very outspoken against the inclusion of servers and data storage equipment in the working plan,” the consumer group said in a statement.

However, there is reason to expect new EU regulation – or the alternative, industry voluntary agreements – on servers in the near future.

For this to happen, however, resources must move to something that is more commensurate with the resources that are available in the United States and elsewhere, writes an EU-commissioned study on the evaluation of the effectiveness of the directive by the UK-based Centre for Strategy and Evaluation Services (CSES).

“If the extension of the directive is not to be an empty gesture, it should be ensured that implementation and enforcement of legal requirements is feasible, practicable and cost-effective,” the study says.

POSITIONS:
“We’ve taken major strides in improving data centre energy efficiency through the EU Code of Conduct for Data Centres, the Green Grid and vigorous competition in the marketplace. This is recognised in the Commission’s working plan study,” said Kirsty Macdonald, senior energy policy manager at Intel Corporation

Commenting on the delays in including new products in the Ecodesign Directive, despite the scientific evidence on the table, European Consumers Organisation BEUC said in a statement:

“Another example of the failure of Ecodesign to tackle ICT products are computers. Although they were already in the 2005 transitional list of products, industry has successfully prevented the adoption of any mandatory requirements. And the Voluntary Agreements on imaging equipment and TV-decoders concluded by the ICT industry under article 8 of the Ecodesign Directive have been nothing more than a series of delays and a disappointing list of exemptions.”

Monique Goyens, director general of BEUC, said:

“The proposal to include maximum up to nine products in the Ecodesign Working Plan is a clear indication of a lack of ambition on behalf of the Commission. With scientific evidence in favour of tackling more products than currently proposed, we regret the Commission is shying away from taking bolder steps. We ask the Commission to dedicate the necessary resources to give a powerful impulse to Ecodesign.”

A study on the evaluation of the effectiveness of the eirective by the UK-based Centre for Strategy and Evaluation Services (CSES) says: “The inadequacy of Commission resources for participation in implementing processes is clearly a major cause of delay and a very significant constraint on the whole Ecodesign system. By way of comparison, staffing levels in the USA are in the region of 10 times the number of desk officers in the Commission. Even in China, there are about 70 staff and more than 40 product regulations. There is a similar disparity in terms of resources devoted to the necessary studies.”

Source:http://www.euractiv.com/specialreport-ict-fuelling-economy/eu-eyes-energy-efficient-design-computer-servers-news-510890

Comments Off »

Posted in Hardware News

Tags:

£4bn framework launched for hardware from tablets to servers and storage

February 7th, 2012
Share

The Government Procurement Service has advertised for suppliers to join a wide-ranging £4bn ICT framework.

The framework will be open to public sector organisations for two years, according to a notice in the Official Journal of the European Union, and covers the following lots:

• Desktop client devices: which will include desktop computers, keyboards, mice and computer memory. The GPS says it expects three suppliers to be awarded agreements.

• Laptops equipment: including notebook devices, port replicators/docking stations, and associated equipment, for which four suppliers will be signed up.

• Tablet/slate devices: five suppliers will be awarded contracts.

• Monitor device equipment: to include wall brackets for monitors; desk stands for monitors and speakers, and three contractors are expected to be signed up.

• Thin client devices: contract awarded to three companies.

• Servers: to include tower, rack and blade servers, server chassis/standard racks, power supply units, server hard disks, hard disk arrays and server memory. Three suppliers will be signed up.

• Storage devices: delivered by three suppliers.

• Network switch devices: delivered by three suppliers.

• Desktop printers: to include printer memory, paper trays and power cables and delivered by five suppliers.

• ICT peripherals: awarded to three suppliers.

• Non-standard products related to desktop hardware, services and solutions, which will be awarded to five suppliers.

• Non-standard infrastructure hardware, services and solutions, for which eight companies are sought.

The framework will be open to include central government departments and their arm’s length bodies and agencies, non-departmental public bodies, NHS organisations and local councils.

Source:http://www.guardian.co.uk/government-computing-network/2012/feb/06/gps-four-billion-ict-framework?newsfeed=true

Comments Off »

Posted in Hardware News

Tags:

SeaMicro Bets On Bigger Chips for ‘Micro Servers’

February 1st, 2012
Share

Reducing power and space is a pressing issue in corporate computer rooms, with some hardware makers adopting chips used in mobile devices to build what the industry is calling “micro servers.” SeaMicro, which helped invent the category, just broadened it.

The Silicon Valley startup on Tuesday announced plans to begin offering new machines that use a variety of Intel’s Xeon chips for servers, where its other systems use the Atom chips found in portable PCs called netbooks. That shift, the company says, allows customers to use its machines for an array of new tasks that could not be addressed before, such as running Web databases rather than simply serving up Web pages.

SeaMicro
SeaMicro’s latest “micro server”

SeaMicro’s new $138,000 machines boasts 64 four-processor Xeon chips along with the startup’s proprietary chips for connecting them together and reducing energy consumption. They use half the energy of comparable arrays of servers, fit in about a third the space, and have 12 times the communications bandwidth between the systems, said Andrew Feldman, the company’s chief executive.

SeaMicro unveiled the system at a press event with Intel and Samsung Electronics, which is supplying memory chips for SeaMicro systems. Executives from the two companies praised the new server design lavishly.

But asked just how big the micro-server category could become, Intel data-center unit general manager Jason Waxman put the potential at only about 10% of the server market. That’s a lucrative niche, to be sure, but Intel may be being a bit conservative, say some analysts, like Jean Bozman of IDC.

“Why couldn’t it be 30 or 40%?” asks Roger Kay, an analyst at Endpoint Technologies Associates.

One reason for Intel’s view, Kay theorizes, is that some jobs depend less on ultra-fast communications than SeaMicro’s systems excel at and more at heavy duty processing power. But another may have to do with price of SeaMicro’s systems and Intel chips, and the threat of cannibalization; the power-efficient Xeon version SeaMicro is using is priced at less than $300, where Intel sells some Xeon chips for thousands of dollars.

Either way, Feldman says it is selling its existing machines like hotcakes. He dodged a question about whether the company has plans to go public, but said “business has been phenomenal,” with first-year revenues surpassing four well-known recent startups combined.

Feldman noted that the company is getting particularly good traction with military and other government customers, some of whom are deploying its servers in the field to analyze data locally before sending it back to headquarters. “They are taking these machines to extraordinarily unorthodox places,” he said.

Intel’s chips are not the only option for companies like SeaMicro. Rival Calxeda, for examples, has talked up plans for micro servers using chips from ARM Holdings.

Feldman stressed that SeaMicro does not rule out eventually trying chip architectures, too. “We expect ARM to be formidable competitors,” Waxman of Intel said. “It’s up to us to build the best engine.”

Source:http://blogs.wsj.com/digits/2012/01/31/seamicro-bets-on-bigger-chips-for-micro-servers/

Comments Off »

Posted in Hardware News

Tags:

New Opteron Servers Struggle Badly Despite Additional Cores

November 25th, 2011
Share

In the wake of Bulldozer’s weak desktop launch, there was some hope that server benchmarks would redeem the chip and give AMD a fighting chance against Intel in at least one of its three core markets. Recent reviews and a comprehensive roundup of formally published results, however, indicate that while BD fares a bit better in servers, it’s far from vindicated.

AMD’s new 6200 Opterons (Interlagos) are often a sidewise move from the older 6100 series, which is based on Magny-Cours. Interlagos is essentially trapped by the same trio of issues that cripple Bulldozer. The shared core design hurts SMP scaling; a 16-core Interlagos scales more like a 13-14 core Magny-Cours. The clock speed increases are offset by Bulldozer’s lower single-thread performance. Finally, Interlagos draws more power than its predecessor, which in turn prevents further clock speed scaling.

As Ars Technica notes, Interlagos is faster than its predecessor in some tests — but often by much less than one might expect. In JBB2005, Interlagos is just 27 percent faster than Magny-Cours, despite 33% additional cores and a four percent clock speed advantage.

If AMD had even managed to maintain single-thread performance parity with Thuban, than the chip’s higher clock speeds and greater number of cores would give it at least a marginal fighting chance, particularly in highly multithreaded environments. As Ars writes, “AMD compromised single-threaded performance in order to allow Bulldozer to run more threads concurrently, and that trade-off simply hasn’t been worth it… it leaves us wondering what might have happened if AMD had simply extended its old architecture. Another four cores in a Magny-Cours processor would show close to the same 33 percent gain, and would do so without compromising single-threaded performance.”

The same thought has occurred to us. Bulldozer/Interlagos was supposed to save die space compared to Magny-Cours, but AMD’s decision to load the chip with 16MB of L2+L3 cache per eight-core configuration largely obviated the space savings. At this point, extending Thuban does look as though it would’ve been the smarter option.

The problem with that argument, however, is that it wouldn’t have addressed the greater issue. While it’s true that a 16-core Magny-Cours and subsequent 8-core 32nm desktop chips would’ve been more attractive than Bulldozer, they wouldn’t have improved AMD’s competitive position vs. Intel. Sunnyvale would still need a new CPU architecture that was capable of competing with Sandy Bridge, and it still wouldn’t have one.

There’s no quick fix for Interlagos, but there’s still the possibility that future iterations of the chip will prove more capable. Right now, that — and a handful of tepid server wins — is the best AMD has to offer.

Source:http://hothardware.com/News/New-Opteron-Servers-Struggle-Badly-Despite-Additional-Cores/

Comments Off »

Posted in Hardware News

Tags:

Japan Pushes World’s Fastest Computer Past 10 Petaflop Barrier

November 3rd, 2011
Share

Japanese IT giant Fujitsu and the government-funded RIKEN research lab announced that the supercomputer they’ve built in Kobe can speed through 10.51 quadrillion floating point operations per second.

Known as the “K Computer,” this is the first megamachine to achieve 10 quadrillion operations a second — aka 10 petaflops — and it will be named the world’s fastest supercomputer when the official Top 500 list is unveiled next week.

Earlier this week, the Chinese made headlines when they uncloaked a world-class supercomputer built entirely with homegrown processors, but this “Sunway” cluster is unlikely to crack the global top ten. The K Computer is ten times faster — when it revs up to peak performance.

When the last Top 500 list was announced this past June, the K Computer took home the ultimate prize, and earlier this week, Jack Dongarra — the University of Tennessee professor who oversees the list — told us the Japanese machine will retain its crown when he releases the latest list at the SC11 trade show in Seattle on Nov. 14.

The K Computer was commissioned by the Japanese Ministry of Education, Culture, Sports, Science, and Technology, with the aim of breaking the 10 petaflop barrier. The K is named for the Japanese word “kei”, which represents 10 quadrillion. The original idea was for indigenous IT giants Fujitsu, Hitachi, and NEC to collaborate on the $1.2bn project, but in the wake of the 2009 worldwide financial crisis, both NEC and Hitachi dropped out of this all-Japanese effort.

This left Fujitsu to build the cluster using its own SPARC64 VIIIfx processors — chips specifically designed for this sort of “high performance computing” (HPC). The K Computer spans 864 server racks and over 88,000 interconnected CPUs, all able to work towards a common task. The last of the racks were installed at the RIKEN lab in Kobe, Japan this past August, and now, Fujitsu has released the official benchmarks — a hallowed moment in the world of supercomputing.

According to the industry standard Linpack benchmark, the K Computer’s average performance is about 93 percent of its peak 10 petaflop speed. In June, it topped the Top 500 list with a peak performance of 8.162 petaflops.

But it won’t keep its crown for long. The kings of the supercomputing game never do. IBM and Cray are building 20 petaflop machines for the Department of Energy, and these should go live next year.

This morning, during a panel discussion at the Four Seasons Hotel in San Francisco, several HPC gurus discussed the next major milestone: the exascale supercomputer, a machine that’s one hundred times faster than the K. But this is likely several years away.

“We’ll see hardware capable of producing exascale calculations in 2017 or 2018,” said Margaret Williams, senior vice president of HPC Systems at Cray. “But will we see software applications that can actually do sustained performance at that level? I think that won’t be until at least 2020.”

Theoretically, we could build an exascale supercomputer today, but it wouldn’t be financially viable — mainly because it would use far too much power. The aim, said Chuck Moore, a corporate fellow and technology group CTO at chip-designer Advanced Micro Devices, is to achieve exascale performance at 20 megawatts of power. Each megawatt, he said, translates to about a million dollars a year in electricity costs alone.

Fujitsu was not able to give us the official power consumption of the completed K cluster, but in June, when it reached a one petaflop peak, it consumed 9.89 megawatts — aka $9.89 million dollars a year.

Source:http://www.wired.com/wiredenterprise/2011/11/japanese_megamachine/

Comments Off »

Posted in Hardware News

Tags:

Facebook building customised servers: Dell, HP losing

September 12th, 2011
Share

When Facebook Inc set out to build two new data centers, engineers couldn’t find the server computers they wanted from Dell Inc or Hewlett-Packard Co. They decided to build their own.

“We weren’t able to get exactly what we wanted,” Frank Frankovsky, Facebook’s director of hardware design, said at a conference on data-center technology last month.

Hewlett-Packard, Dell and companies that sell the computers off the shelf are losing sales in a key market because Facebook and larger rival Google Inc are leading a switch among Internet companies to do-it-yourself servers. These customized machines now account for 20 per cent of the US market for servers, which generated $31.9 billion globally in last year, said Jeffrey Hewitt, an analyst at Stamford, Connecticut-based Gartner Inc.

As sales of personal computers slump and consumers shift to tablets such as Apple Inc’s iPad, computer makers are becoming more dependent on servers. Dell and Hewlett-Packard lose out when they’re shunned by large customers such as Facebook, which are outfitting data centers with thousands of servers.

“It’s definitely a threat to the traditional business model,” said Jim McGregor, chief technology strategist for researcher In-Stat in Scottsdale, Arizona. “Customers are finding solutions that the industry wasn’t ready to provide.”

Buyers say custom servers provide a cheaper, more efficient way of meeting the boom in demand for personal data shared via the Web. A lot of that demand can be met by less expensive machines shorn of the components, upgrades and backup services that server makers traditionally offer to large corporations.

Hold the ‘pickles’
“People want to be able to build it their way,” Frankovsky said at the Dell-Samsung Chief Information Officer Forum in Half Moon Bay, California. “They kind of want a Burger King: ‘I don’t like pickles — why do I have to have pickles?’”

The market for servers, which increased almost 20 per cent in the second quarter, is outpacing growth in desktop PCs as companies upgrade their corporate networks and snap up the thousands of servers needed to run data centers.

That traditionally has spurred demand for machines made by Hewlett-Packard, Dell and International Business Machines Corp. Yet surging costs for running and maintaining data centers is causing Facebook and other companies to seek cheaper options.

Hewlett-Packard, which last month cut its profit forecast for the third time since November, can’t afford to lose momentum in one of its better-performing units. Hewlett-Packard’s revenue from the servers that are typically deployed in so-called cloud-computing data centers rose slower than the industry average in the second quarter, according to Gartner.

Slow growth
Dell, where sales have barely budged for two quarters, needs to keep server customers happy too. Dell’s sales of servers based on PC chips — the type most often used in cloud-computing data centers — grew 4.4 per cent in the June period, according to Gartner. Cloud-computing networks store and deliver software and services via the Internet.

In one indication of the growing demand for servers that are being built from the ground up, Intel Corp said its revenue from chips used to craft servers for data centers surged 50 per cent in the second quarter.

Intel’s understanding that cloud data centers had different needs was key to a turnaround in market share. The company share of the cloud server-chip market is above 90 per cent from a low of 35 per cent in 2007.

‘Different animal’
“It’s a completely different animal” than corporate servers, said Rejeanne Skillern, head of marketing at Intel’s cloud computing division.

Google, Facebook and Microsoft Corp have designed servers that contain the minimum amount of components required for their specific task.

Facebook’s servers, for example, have custom power supplies and circuit boards in sheet-metal enclosures designed to maximize airflow with the minimum number of fans.

Those and other tweaks — combined with a specially designed facility — boosted efficiency by 38 per cent and reduced the cost of building a data center in Oregon by 24 per cent, according to the company.

Google’s servers are also built to the company’s specifications, with hardware limited to what is needed for applications to run. The machines run a stripped-down version of the Linux operating system that leaves out unnecessary code.

Microsoft is hedging its bets. The software maker designs its own servers. Yet it also has big computer makers build them, sticking with suppliers that can provide worldwide support to quickly fix any server that breaks down, said Dileep Bhandarkar, chief architect of Microsoft’s Global Foundation Services.

Custom designs
For the data centers that underpin its cloud services and Bing search engine, Microsoft uses machines similar to the scaled-down, low-power ones Google uses. Though instead of making the hardware itself or through contractors, the company has a team of engineers who create server designs. Microsoft then commissions companies like Dell and Hewlett-Packard to build the machines.

Computers makers must come up with products that fit the new needs of data center builders.

“Many of them are realizing they’re going to have to address that,” said Gartner’s Hewitt, who is based in San Jose, California. “Most of this is being driven at the moment by Google, but there are others who will look at it.”

Microsoft and other companies’ demands for tailor-made machines leave server makers with a dilemma: go after build-to-order market and accept lower profits, or lose customers and relevance.

‘Execution team’
Hewlett-Packard sees the market as an avenue for growth, said Jim Ganthier, vice president for marketing in the company’s server unit.

“We’ve gotten entire design teams, we’ve got an entire execution team that wakes up every morning thinking about this,” he said. “If you think you’re going to go in there with a generic off-the-shelf product, you’re going to lose.”

Dell, meanwhile, supplies customized servers to Microsoft and Facebook, as well as Chinese Internet companies Tencent Holdings Ltd and Baidu Inc, said Tim Mattox, vice president of worldwide enterprise product management.

“There’s quite a lot of opportunity, but there’s not a lot of margin in it,” said Matthew Eastwood, an analyst at Framingham, Massachusetts-based IDC.

The computer industry has little choice but to adapt. The shift to using cheaper servers for cloud computing may spread beyond Internet search and social media and find its way into other areas, such as mobile-phone service providers.

Source:http://timesofindia.indiatimes.com/tech/enterprise-it/infrastructure/Facebook-building-customised-servers-Dell-HP-losing/articleshow/9954260.cms

Comments Off »

Posted in Hardware News

Tags:

Intel Unveils 10-Core Xeons, Mission-Critical Servers

April 6th, 2011
Share

Intel announced its new E-series of Xeon processors today, claiming that the new processors will deliver nearly unparalleled advances in CPU performance and power efficiency. It’s been just over a year since Santa Clara released its Nehalem-based octal-core Beckton processors. Whereas Beckton was focused entirely on performance and architectural efficiency, these new Xeons are more balanced. The new chips boost the core count to ten (up to 20 threads with HT enabled) and will be offered at a wide range of TDPs.

ntel’s presentation made it clear that it’s gunning for what’s left of the RISC market. Kirk Skaugen, vice president and general manager of Intel’s Data Center Group, made a point of telling the conference that there’s “No workload in the world today that Xeon can’t handle.” History certainly favors his words. Intel’s quoted figures indicate that while the high end of the server market grew just five percent from 2002-2010, Intel’s share of it nearly doubled.

The company went out of its way to note that Itanium’s share of the market grew enormously over the past eight years, but it’s Xeon, not Poulson, Intel is betting on. The new E7 series incorporates the benefits of Sandy Bridge, its support for new instructions, and its improved power management technology. Intel has also baked in support for low-voltage DIMMs, which allows vendors to opt for 1.35v products. The power savings, at 1W per DIMM, might not sound like much, but the E7 series supports up to 2TB of RAM in a 4S system. According to Intel, low-voltage DDR3 can cut a server’s power consumption by up to 128W.

OEM support for the new E7 processors seems downright enthusiastic; 19 vendors have announced a total of 35 systems with shipping to begin immediately. This may be partly due to the way the E7 helps to simplify Intel’s product mix. Up to now, Intel’s heavy-hitting Beckton was a 45nm chip that lacked the 32nm enhancements of the Xeon 5600 parts.

Source:http://hothardware.com/Reviews/Intel-Unveils-10Core-Xeons-MissionCritical-Servers/

Comments Off »

Posted in Hardware News

Tags:

Get Adobe Flash playerPlugin by wpburn.com wordpress themes