According to Bloomberg, the deal would see IBM’s low-end x86 server business be bought out by Lenovo, and it could cost as much as $4.5 billion, with $2.5 billion being on the lower end. IBM’s server business hasn’t been doing so well, so a selloff to Lenovo would mostly likely put IBM’s server business in better hands.

For example, Lenovo used the purchase of IBM’s PC business to become the world’s second-largest PC manufacturer, and now the company is dabbling around in the tablet market, as well as the smartphone market for the first time. Lenovo told its investors today they’re “in preliminary negotiations with a third party in connection with a potential acquisition,” but IBM was not mentioned.

IBM’s latest quarterly earnings weren’t terrible, but they also weren’t good. The company reported a 1% drop in profits over the quarter, which missed predictions by $1.3 billion. Furthermore, IBM’s System x server sales dropped by 9%, while mainframe sales increased 7%. However, IBM’s hardware profits experienced a $405 million loss.

[via Bloomberg]

The company mentioned last week that they would be announcing the next generation in Moonshot hardware, and as expected, they’re starting out the week on a high note with the new line of servers. HP mentions that with the growing needs of internet users, cloud data storage will grow exponentially, to the point where larger and larger facilities will have to be built in order to meet the demand of storage needs.

However, HP says that Moonshot will dissolve the need for bigger and bigger data centers, saying that the new Moonshot servers can take up 80% less space, as well as use up to 89% less energy and be 77% less costly than traditional servers. HP even went as far as saying that the new Moonshot systems are comparable to the move from Unix to x86 servers, as well as the innovation of blade servers.

The new Moonshot systems are available today in the US and Canada, while availability in Europe, Asia, and Latin America will start next month. Pricing begins at a whopping $61,875 for the enclosure, 45 HP ProLiant Moonshot servers, and an integrated switch, but clients will be able to customize their setup to fit their own needs.

In a blog post, EA detailed some of the changes and upgrades they made throughout the week, including the completion of migrating all of their original servers to the new, faster and higher capacity servers. However, while the migration is complete, EA says that they’ll continue to provide “ongoing optimizations” for the servers.

EA also launched a server status page, so that gamers could see the status of server operations from around the world. EA says these are updated live. The page will list whether a server is available for play, or if its full. Servers that are experiencing high volume will be marked with yellow. However, most of the servers are currently green at the moment, which means they’re available.

Last but not least, EA released patch 1.5, which addresses “some trade intermittency issues.”EA says this should make regional interaction more responsive, and the publisher is going to continue to improve the speed between cities within a region. The patch should also decrease the chance of losing progress on your city via a rollback.

These music servers seamlessly integrate with a handful of popular AV Receivers, including those from Harman Kardon, Onkyo, Pioneer, Sony, and Yamaha. The CasaTunes Music Servers start out at three wired rooms and five wireless rooms, but they support up to a whopping 24 wired rooms and 10 wireless rooms, making it the perfect solution if you happen to own the Palace of Versailles.

Of course, the system allows users to install a wired-only music system, a wireless-only system, or a hybrid of the two in any combination, which makes it fairly versatile as far as where the system can be implemented. And the technology uses Apple’s AirPlay protocol to beam music wirelessly to various receivers.

Users can also listen to different playlists in different rooms, and it can all be controlled via the optional keypad (pictured above), through a web browser, or using the CasaTunes mobile app for iOS and Android. While setting up a multi-room music system isn’t cheap by any means, it’s a great way to get audio blasting all throughout your house, especially during a party.

The Pirate Bay’s announcement about moving their servers to North Korea existed right up on their homepage’s front page. From there they showed a lovely North Korean set of sails on their own branded pirate ship, proceeding to tell the world that North Korea – of all places – invited them to host their servers and become a new beacon of freedom. But it was not to be.

Down on The Pirate Bay’s Facebook page, the team has shown the world that they’re “in it for the lulz”. Speaking for TPB, users Winona, Winnie, and Winston Bay have spoken up about how creating a ruse such as the one they’ve executed this week proves that they “don’t run TPB to make money.” They also call out the RIAA (cleverly re-named MAFIAA in the statement) for being a “profit hungry idiot”.

“We’ve hopefully made clear (once again) that we don’t run TPB to make money. A profit hungry idiot (points at MAFIAA with a retractable baton) doesn’t tell the world that they have partnered with the most hated dictatorship in the world. We can play that stunt though, cause we’re still only in it for the f***** lulz and it doesn’t matter to us if thousands of users disband the ship.” – TPB

Of course the fun didn’t end there, the team continuing by calling out none other than the readers and the users – all of us, that is – for not calling them out when they announced such an absurd notion as moving to North Korea.

“We’ve also learned that many of you need to be more critical. Even towards us. You can’t seriously cheer the “fact” that we moved our servers to bloody North Korea. Applauds to you who told us to f*** off. Always stay critical. Towards everyone!” – TPB

So that pins it, in the end. We now live in a world where we cannot trust any company to speak the truth about their own organization – or it may just be that we’re not going to be able to read anything the team at The Pirate Bay publish from this point forward without a large grain of salt. That sounds reasonable – as does remembering that anyone could make a mistake in this fast-paced online publishing world of ours.

How do you feel about the situation?

According to AMD, it’s not only low-power mobile chips but next-gen server processors that will benefit from the fresh blood.  Matar and Meretsky bring “years of expertise in SoC design and developing 64-bit software ecosystems, respectively,” Mark Papermaster, AMD’s Chief Technology Officer, said in a statement today. “The fact that these computing experts have returned to the company underscores AMD’s unique position and opportunity, based on differentiated IP, to take a leadership position in low-power clients and dense cloud servers.”

Matar will take up a role as corporate vice president of SoC Development, where AMD says “he will employ his in-depth understanding of market trends and wireless industry requirements to help AMD shift appropriate focus to client and embedded market opportunities.” The exec has previously worked at Motorola and various semiconductor companies, before his stint helping with low-power chip designs at Qualcomm.

As for Meretsky, he will be corporate vice president of Software IP Development, and will bring his code expertise honed from Apple and, before that, P.A. Semi to work on AMD’s software development activities. His role will be in “ensuring tools will be in place that will enable developers to take advantage of the compute power in AMD’s SoC designs.”

“Meretsky is a former AMD Fellow” the company points out, “who was instrumental in making AMD’s 64-bit technology the de facto industry standard for PCs, workstations and servers.” It also highlights that Meretsky worked alongside Jim Keller back in the P.A. Semi days; Keller was snatched by AMD from Apple, where he was chip architect, last year.

AVADirect also partnered up with ASUS and used the company’s ESC4000 G2 Black 2U server platform to bring some respectable hardware to these new server units. On the inside, you can store up to 512GB of DDR3 RAM, with room for eight PCI-E 3.0 16x expansion cards (including those four Tesla units).

You’ll also find eight SATA ports with multiple RAID configurations, eight hot-swappable hard drive bays, and a 1+1 Redundant 1,620W 80PLUS Platinum power supply. That’s certainly a lot of firepower that must be treated carefully. AVADirect says that these new units are meant for render farms, imagine generators, rack-mounted graphics workstations, and multi-display workstations.

Of course, AVADirect will let you customize the new server system to fit your needs, and their website offers a ton of different configurations. However, the company’s recommended build costs just shy of $5,250, so if you’re serious about getting some major computing power in your hands, you better be saving up right this instant.

Intel says that there are already over 20 low-power designs including micro servers, storage, networking systems, and others that use the Intel Atom processor S1200 family. The processor family promises server class levels of reliability and manageability while saving on costs. The system-on-a-chip includes two physical cores and a total of four threads enabled via Intel hyper-threading technology.

The system offers 64-bit support, a memory controller supporting up to 8 GB of DDR3 memory, support for Intel Virtualization Technologies, ECC, eight lanes of PCI Express 2.0, and other I/O interfaces integrated from Intel chipsets. Intel will offer the processor family with three processors ranging from 1.6 GHz to 2.0 GHz.

The new processor family is compatible with the x86 software commonly used in data centers around the world right now. The hallmark feature of the product family is that the it is the first low-power SoC supporting server-class features with as low as 6 W of TDP. The processor family is shipping today with pricing starting at $54 in quantities of 1000 units.

AMD also says that the video card offers 5.19 Tflop of single-precision and 1.48 Tflop of double-precision floating-point calculations. The video card uses AMD’s next-generation Graphics Core Next Architecture. The powerful graphics card is aimed at use in a variety of fields including finance, oil exploration, aeronautics, automotive design and engineering, geophysics, life sciences, medicine, and defense.

The card offers dual GPUs and has high throughput and low latency transfers for quick computing of complex calculations that need high accuracy. The video card has 6GB of GDDR5 RAM and a 384-bit interface. Output options include four mini DisplayPort outputs.

The card also has one standard DVI output. Maximum resolution supported on the card’s DisplayPort 1.2 port is 4096 x 2160 with standard resolution on other outputs being 2560 x 1600. The video card consumes 375 W of power and needs to slots inside a computer. The retail price on the card is $3599.

The Opteron 6300 Series line-up maxes out at 3.5GHz base frequency, though there’s up to 3.8GHz on offer in AMD Turbo CORE mode. 4-, 6-, 8-, 12-, and 16-core versions are offered, with TDPs ranging from 85W in the 6366 HE low-power model, through to 140W for the 16-core, 2.8GHz 6386 S top of the line chip.

Up to four 1866MHz memory channels are supported, and AMD claims the 6300 Series is the only x86 processor to work with ultralow voltage 1.25v memory. Each CPU can handle up to 384GB of memory – spread over up to 12 DIMMs – and up to four x16 HyperTransport links (each up to 6.4GT/s).

However, AMD isn’t solely relying on x86 for its future server chip strategy. The company recently confirmed that it was developing 64-bit ARM-based server processors, borrowing architecture more commonly associated with tablets and smartphones, and repurposing it for frugal use in enterprise server rooms.

The first servers to use the Opteron 6300 Series chips are on sale today, with Dell and HP both signed up to produce systems using AMD’s new CPU by the end of the year.

According to the press release, the increase in mobile devices and new business models has put a particular strain on data centers’ server infrastructure. The market, says AMD, is actively seeking ways to increase efficiency while reducing ownership costs. ARM-based processors are ideal for parallelization, making them a better choice over large-core CPUs.

Said ARM CEO Warren East, “The industry needs to continuously innovate across markets to meet customers’ ever-increasing demands, and ARM and our partners are enabling increasingly energy-efficient computing solutions to address these needs. By collaborating with ARM, AMD is able to leverage its extraordinary portfolio of IP, including its AMD Freedom supercompute fabric, with ARM 64-bit processor cores to build solutions that deliver on this demand and transform the industry.” AMD’s president and CEO Rory Read also offered a statement, saying that AMD will lead the industry in widespread adoption of more energy-efficient 64-bit processors.

This comes after AMD’s acquisition of SeaMicro, which was announced back in March. Says the press release, AMD will be integrating the SeaMicro fabric with the upcoming processors to enable more energy-efficient processor clusters. AMD made the announcement earlier today at a hosted event in San Francisco. No details were released on the processors.

[via AMD]

Today’s outage appears to be centered around their EC2 operations with Amazon’s status for the area currently flipped to “performance issues.” This status does not include official word from Amazon as of yet, but the cards do not lie. The EC2 status as listed by Amazon’s own Amazon Web Services status listing have gone from “degraded performance” at 10:30 AM PDT to “new launches for EBS backed instances are failing” closer to 11:30 AM PDT.

You can watch the updates yourself at Amazon’s Web Services update page while you wait for the web giants to come back – meanwhile we’ll be following along and updating you from SlashGear here as it happens as well. The current statuses surrounding North America, especially in the Eastern region of the states, let us know that Amazon is not only seeing outages in several areas, but that re-launching said services is currently a no-go.

If you have a peek at Outage Analyzer you’ll find that they’ve made it clear that Amazon’s outages have likely affected nearly 100 domains across North America up through Canada and Alaska as well as through the UK and surrounding countries. The outage remains open through the 3PM CST hour. This list of downed sites continues to grow with Airbnb, Heroku, and Netflix being added here right before this post is published – stay tuned!

UPDATE: 3:10 PM CST – Affected sites list grows – Pinterest, FastCompany, FourSquare added. Failures appear sporadic as some sites appear to be up and running intermittently while others remain down for the count.

UPDATE: 4:19 PM CST – Several sites (including Reddit) appear to be coming back up for air now here and there – of course heavy loads are expected due to the mass amounts of cat pictures that were waiting to be uploaded and viewed.

Calxeda is developing a chip codenamed Lagos, which will be based on ARM’s 64-bit ARMv8 architecture. ARM announced their 64-bit architecture almost one year ago to the day, and has said themselves that servers with their chips based on its 64-bit architecture would be available around 2014 or later.

Calxeda recently received $55 million in capital funding last week, and the company is planning to use this funding to focus on their chip development. Calxeda currently offers a 32-bit prototype server chip called EnergyCore, which is based on ARM’s Cortex-A9 processor design. It’s used in prototype servers from companies like Dell and HP.

The reason why ARM processors are becoming more and more popular for server use is that they’re way more power-efficient than chips from Intel. And in order to keep up with the growing demands of computing, Calxeda is looking to transition to these energy-efficient 64-bit chips in just a couple of years.

The digital walkthrough is split into three sections – tech, people, and places – and while there isn’t a huge amount of technical detail, the photos are certainly eye-catching. Rack upon rack of caseless ‘boards, with eye-catching loops of colored cabling splashed across them.

Meanwhile, Wired‘s Steven Levy got to go one better than browsing an online gallery, and walked through the hallowed data halls himself. Along the way he picked up a few details, such as how a massive room full of interlinked machines might not necessarily be the best way to serve up information.

“It would be slow and burdensome to have millions of people grabbing videos from Google’s few data centers. So Google installs its own server racks in various outposts of its network—mini data centers, sometimes connected directly to ISPs like Comcast or AT&T—and stuffs them with popular videos. That means that if you stream, say, a Carly Rae Jepsen video, you probably aren’t getting it from Lenoir or the Dalles but from some colo just a few miles from where you are” Steven Levy, Wired

Google’s comments about the sanctity of user data come at a tricky time for the company. EU privacy regulators demanded yesterday that the company modify its privacy policies after the unified T&Cs came into effect earlier this year, arguing that the system was insufficiently clear to individual users as to how their data might be used or shared.

With Synology’s new System Hibernation mode, the DiskStation DS213+ is able to actively listen for network activity while consuming only tiny amounts of power. When it detects network activity, it’s able to quickly resume full operations. In System Hibernation mode, this machine requires just 2.64 watts of energy consumption.

When the device is in full operation mode, this device attains average read speeds of 110.36 MB/sec, while write speeds average at 84.31 MB/sec – both of these numbers are under RAID 1 configuration in a Windows environment. With AES 256-bit encrypted r/w speeds you’ll have 49.6 MB/sec for read and 24.7 MB/sec for write.

The DS213+ includes a a 64bit floating-point unit to boost the speed at which you’re able to decode and encode on the daily. Another bonus here is the DS213+’s upgraded surveillance environment with the ability to stream 12 channels of high definition IP camera all at the same time.

This device also has two USB 3.0 ports, one eSATA port, and a full-sized SD card slot. Inside you’ve got a noise dampening design and a 92 mm system fan for next-level airflow with just 19.9 dB(A) of noise. Like every other Synology NAS Server you’ll be working with DiskStation Manager 4.0 (DSM 4.0), and the whole system is available this week around the world wherever fine Synology products are sold!

[via Synology]

Apparently, someone unintentionally cut through fiber cables that connected Wikipedia’s servers located in Florida to the Internet. Wikipedia denies any foul play in the outage, and a spokesman noted that two overland cables had been severed between Tampa and Virginia. The slashed cables took Wikipedia off-line for over an hour and once the lines were fixed it took another hour to get servers back up and running.

Wikipedia has been having an interesting year. The company took itself off-line as part of a protest against SOPA laws intended to help prevent piracy. Wikipedia’s parent company Wikimedia has suffered from controversy surrounding a key executive in its UK division recently after Ashley Van Haeften resigned as the chairman of Wikimedia UK.

Van Haeften stepped down in the middle of a controversy over personally posting links to pornography in the biography of a living person. As a response, he was banned indefinitely from contributing to the English version of Wikipedia by ArbCom, which is elected committee of senior Wikipedia editors. ArbCom also found that he had violated editing rules by using multiple accounts to change pages.

[via Telegraph]

That’s a total of eight Kepler cores all running in parallel. AVADirect hasn’t shared the rest of the specifications of the server itself, nor benchmarks – which is, we must admit, what we’re particularly keen to see.

Why would you want four high-power graphics card in a server? Well, while gaming probably isn’t high on the agenda, turning NVIDIA’s CUDA cores into a parallel processing workstation could have some significant benefits for anyone doing graphics or 3D rendering, or crunching huge quantities of mathmatical data.

NVIDIA normally pushes its Quadro or Tesla cards for dedicated parallel processing tasks, but there’s no reason the eminently capable GTX 690 – which has 3,072 CUDA cores apiece – shouldn’t turn its hand to something more serious than Crysis. No word on overall system pricing, but with each of those EVGA GeForce GTX 690 4GB cards coming in at nearly $1,100 you’re looking at almost $4,400 for CUDA cores alone.

This set of versions starts with Foundation, that being an economical general purpose server aimed bit of software with pricing only revealed to manufacturers of the devices it’ll be pushed to. Next you’ve got an Essentials version of Windows Server, this edition made for small business environments and costing $425. Pricing on each of these editions, mind you, represents Open No Level (NL) ERP.

The Standard edition of Windows Server 2012 is aimed at low density or non-virtualized environments and is one of two editions to be made for Processor + CAL licensing models, here priced at $882 USD. The other of these two editions is the Datacenter edition which is aimed at highly virtualized private and hybrid cloud environments. The Datacenter edition has unlimited virtual instances and will cost you $4,809.

You can also try the software out before you drop hundreds or thousands of dollars for your business right over at the Windows Server 2012 Release Candidate site right this second!

Check out the timeline below to dive deep into more Windows Server action and let us know if your business will be upgrading this year!

App Engine currently sees 7.5bn hits per day, and two trillion DataStore operations per month. Google Compute Engine throws in the scale and experience Google has built up over the years, together with the connectivity of the company’s mobile backbone.

Google’s primary example of the power of Compute Engine was a DNA crunching Institute for Systems Biology program. The center had been using a 1,000 core cluster that produced a single result every ten minutes; with Google Compute Engine’s 10,000 cores, it punched out a result every few seconds. With 600,000 cores, there were multiple results every second.

As for pricing, Google says Compute Engine will offer up to 50-percent more processing, per dollar, than the company’s rivals. Access is in limited availability from today, with up to 10,000 cores for systems that have heavy I/O requirements, while those which have lighter I/O needs can take advantage of hundreds of thousands of cores – Google added 771,886 alone during the presentation.

HP say that the new system using Centerton instead of Calxeda is “an ideal solution for web serving, offline analytics and hosting.” Interestingly, HP make a point of saying that Intel’s solution should be just as cost effective and power efficient as the previous ARM-based system. HP and Intel will also be working together on future generations of Atom processors, further reducing power consumption.

The Gemini system is currently being using in HP’s Discovery Lab in Houston, Texas, but will be made available to customers for testing “soon.” A production version will ship towards the end of the year.

It’s an interesting move from HP, and an important win for Intel. ARM pulled off quite a feat by coaxing HP to use its chips for better server scaling, but it looks like Intel isn’t going to let ARM into the enterprise and server industry without a fight. Intel has been trying to maximize Atom’s performance while reducing power for quite some time, to the point where one version of the chip is now included in smartphones like the Orange San Diego and Lava Xolo X900.

Update: HP has been in touch to clarify that ARM chips will still be used for the Redstone development platform, which is “still on track to begin shipping to HP Labs and select customers later this year.” Gemini, this initial server, is “processor neutral” HP says, and there will be ARM-based versions eventually.

Each blade contains four ARM processors with a matching SATA drive each, making a fully-loaded Copper chassis capable of running 48 hard drives or SSDs on just 750 watts – about as much power as a souped-up gaming computer. Each blade gets a single RAM DIMM supporting up to 8GB and one Ethernet port. The servers are being sent out to select Dell customers in a seed program, so general availability won’t be coming for some time.

Desktop operating systems for ARM are still a little thin on the ground, so Dell will be shipping the systems with Canonical’s Ubuntu build tailored for ARM processors. Red Hat is already working on a compatible version of its Linux derivative. While Microsoft is working on Windows 8 RT for ARM hardware, there’s been no indication that a server version is coming any time soon.

[via The Inquirer]

Microsoft ditched Windows Server 8 in place of Windows Server 2012. The software giant is also committed to launching the new server OS this year. The operating system is expected to launch at the same time Windows 8 from normal desktops and laptops launches.

Microsoft has yet to offer a specific timeframe for the launch, but it is widely expected to happen in October of 2012. Windows Server 2012 is available in beta right now. There hasn’t been any indication from Microsoft on when a release candidate version of the server OS might surface.

[via ComputerWorld]

Fusion-io is hoping to target the entertainment industry, particularly those rendering special effects, thanks to the extremely high read and write rates of the SSD. The company says that IOPS are “pretty much irrelevant”, instead believing that bandwidth and access to data is crucial. The ioFX is capable of reading data at up to 1.4GB/s, and write at around 700MB/s.

The card only comes in one size, though: 420GB. Latency is low to ensure fast access to stored data: .068ms for reads, and 15ms for writes. The PCIe card also has a cooling fan in order to cope with the extra heat found within workstations with a lot of horsepower.

UPDATE: This article originally mistakenly had 68ms for reads. This is of course a very large value – the correct number is .068ms.

The price of the ioFX? $2,495, and that includes one year of support for the card. In comparison, the next step up, the ioDrive 2, costs $5,950. The card is compatible with Windows, Mac OS, and Linux.

[via ComputerWorld]

An additional 750,000 people have had their personal information stolen by hackers according to Utah state health officials. The officials note that the hackers downloaded thousands more files than the initially believed. In all the officials now believe 900,000 people had the personal data stolen from health department servers.

The information for the additional 750,000 people includes Social Security numbers according to a health department spokesman, and some of the files also contain information required to verify Medicaid coverage. That information includes names, addresses, and other personal details. The server that information was stolen from was improperly configured allowing the hackers access. The attack resulted in the loss of 224,000 files, some of which contain hundreds of individual records.

[via Washington Post]

Out of those 181,604 hacked records, 25,096 of those apparently had their Social Security numbers compromised as well. The agencies involved in the hack are cooperating with law enforcement in an investigation. So far the investigation has determined that the hackers were able to access the server on March 30 of this year and are located in Eastern Europe. The hackers made off with 24,000 files in the attack. Each one of those individual files can potentially have claims information on hundreds of individuals adding up to the 181,604 clients.

Information on claims stored on the server can include the name, address, birth date, social security number, physician name, national provider identifiers, addresses, tax identification numbers, and the procedure codes that go along with medical billing. The medical claims information had been moved to the new server and that server had a configuration error at the authentication level that allowed the hackers to get around the security systems. The Utah Department of Technology Services has implemented new security measures and is reviewing every server in the state to be sure they are secure.

[via ZDNet]

Samsung in South Korea has sent out confirmation that Patla is now a vice president at the company, but would not comment on the possibility that they’d soon be making server-ready chips for the market. This move may be one that will bring Samsung out of its 2nd place spot for semiconductor revenue, a spot long held by Intel. Other competitors in the wide world of chips include such fighters as Qualcomm, Texas Instruments, and NVIDIA, each of these currently also fighting for sway over the mobile SoC world where Samsung’s chips sit in not only their own devices, but the iPad and the iPhone as well.

If indeed Samsung does move into server chips, it will mean just as big an expansion as was announced earlier this week in their display business. Samsung has done well for itself in the past handful of years in the mobile business as well as in memory chips and displays, so a move to chips for servers only seems natural at this point. Would you trust a Samsung chip in your server this summer?

[via WSJ]

AMD say the new chip is a “server-class product”, but at a desktop price point. They’ve designed the 3200 series to draw a low amount of power, but with the necessary performance to boot. AMD claim that the new chip offers 38% better price performance, and draws 19% less power per core over rival Intel’s Xeon chips.

The nitty gritty details? The 3200 series runs between a 45 and 65W TDP, clocks in at 2.7Ghz that can run up to 3.7Ghz when needed, and supports DDR3 memory up to 1866Mhz. There will be 4 and 8-core versions that will both have 8MB of L3 cache.

The Operton 3200 series begins shipping today, with hardware partners including MSI, Tyan, and Dell. Fujitsu will also be using the chip in their new server, the 3000 Series MX130.

Wired reports that Google is so fearful of spies getting a glimpse at its hardware that it’s handing out lights that workers wear on their heads to work in the dark server racks. Google is leasing the data center space from a company called Equinix, and the reason for the paranoia is at least somewhat founded considering some of Google’s biggest competitors use the exact same data centers. Google’s competitors could be mere racks away.

According to a guy named Chris Sharp, Google removed all the light bulbs from its data center hardware cages at the Equinix data center and handed out miner helmets with lights for people going into the cages to work. Google considers its data center prowess as part of its advantage and designs its own servers and networking gear. I guess a certain level of paranoia is understandable.

Sharp said, “They had us turn off all overhead lights too, and their guys put on those helmets with lights you see miners wear. Presumably, they were bringing up custom-built gear they didn’t want anyone else to see.”

[via Wired]

Featuring Intel’s Sandy Bridge-E architecture, the Xeon E5-2600 family can handle a full eight cores per processor and as much as 768GB of system memory. The additional cores along with a smaller 32nm architecture, as opposed to 45nm, will make the processor 80 percent faster and 50 percent more efficient. It will also be the first server processor family to integrate PCI Express 3.0 support and I/O controller directly into the microprocessor.

Intel aims to address the growing server needs of a more connected world, forecasting at least 3 billion connected users with 15 billion connected devices by 2015. The new server chips have already begun shipping to manufacturers, including HP, Dell, IBM, Oracle, and Cisco, which are all expected to announce Xeon-based server platforms today.

Rival AMD recently acquired startup SeaMicro for $334 million to enter this same enterprise server market aimed at cloud computing. The move was seen as a blow to Intel, which had appeared to have formed a tight relationship with SeaMicro. However, VP and general manager of Intel’s Datacenter and Connected Systems Group, Diane Bryant, dismissed that notion today, saying that Intel was “not impressed” with SeaMicro and had declined to purchase the company.

[via ZDNet]

AMD is interested in SeaMicro’s core IP, which has to do with a customized chip that can handle the networking demands of more than 500 chips, all packed into a very tight space. SeaMicro was able to eliminate all but three of the chips on a standard server motherboard, which yielded servers that consumed only a quarter of the power and one sixth of the space of traditional x86 servers. AMD hopes to license the technology to other server vendors.

The acquisition will be a blow to Intel, which had a close partnership with SeaMicro and even developed a special version of its Atom processor for the SeaMicro servers. AMD expects to close the deal in March and will be paying $281 million in cash with the remainder paid in stock.

[via PC World]

Tilera made headlines back in 2009 with its 100-core Tile-GX chip, targeting data indexing, web search, and video search, especially in systems with high concurrent load. As cloud storage and processing has gained popularity – not least because of the rise in data speeds allowing mobile gadgets to rely on remote data crunching and capacity – so the interest in multicore chips of this sort has increased.

According to the company, it’s working with more than 80 customers and has twenty design wins already for the new 16- and 32-core processors. Potential applications include streamed video- and voice-over-IP systems, real-time image and video processing, and more.

Tilera is offering a number of evaluation systems, ranging from an entire computer on a half-sized PCIe card up to a full-featured appliance packing up to 144 cores per 1U box. Unlike ARM-based manycore systems, like HP’s Project Moonshot, Tilera’s system can run all the regular server apps without recoding.

It doesn’t seem so strange when we look at the infrastructure inside the building, you can see straight away what attracted Google to the location. The old paper mill has a massive underground tunnel that runs from inside the building to the Gulf of Finland where it slurps up cold ocean water. That cold ocean water was originally used to cool a steam plant inside the paper mill. Google wants to use that massive tube that runs into the Gulf to grab cold ocean water to help cool the massive array of servers that will be inside the building.

The cold ocean water will be fed to a heat exchanger to cool the hot servers, which would help make this a green facility and keep all the servers inside the building running at a frosty temperature. The underground tunnel to the Gulf, which was the big selling feature for Google, spans 450m and tunnels through solid granite bedrock. Before making the purchase, Google sent robots with cameras through the long tunnel to ensure that it was in good condition. The old paper mill was previously owned by Finnish pulp and paper maker Stora Enso. The company closed the mill in 2008 citing poor profits. Google has previously eyed cold climates to place its data centers so it could use the frosty climate for cooling and reduce the power needed for cooling servers inside the locations.

“When someone tells you we’ve selected the next data center site and it’s a paper mill built back in 1953, your first reaction might be: ‘What the hell are you talking about?,’” says Joe Kava, Google head of datacenter operations and construction. “‘How am I going to make that a data center?’ But we were actually excited to learn that the mill used sea water for cooling…. We wanted to make this as a green a facility as possible, and reusing existing infrastructure is a big part of that.”

[via Wired]

Any information on the exact specifications of the to-be-announced Ivy Bridge models haven’t been released yet, but the Sandy Bridge E5 chips have already known to run the gamut from a 1.8 GHz quad-core processor to 3.1 GHz in eight-core assortments. That’s some serious processor speed right there, no doubt. Competitors beware, here.

If this possibility moves into the realm of reality for Intel, the Xeon lineup could create potential problems for workstations like the upcoming Macbook Pro refreshed updates. It may force Apple and other competing companies into either choosing Sandy Bridge models and going for a slightly aged, but wider range of processors, picking from a limited Ivy Bridge range, or having to go through a divide in the computer lineup between the pair of Intel architectures.

Up to four memory channels – with support for up to 1600MHz memory – are on offer, with up to 12 DIMMs per CPU available for a somewhat ridiculous potential total of 384GB memory per processor catered for. AMD is also pretty keen to point out that the new Opterons are the only x86 chips to support ultra-low voltage 1.25v memory, adding to the platform’s energy frugality.

“The AMD Opteron 3000 series offering is designed for hosting customers who require dedicated servers for their customers. These cloud and web hosting customers appreciate the cost savings associated with a lower cost infrastructure, yet still want to deploy a server-class product with reliability and security features and server OS certification” AMD

As for what’s in the pipeline, first to market will be the Bulldozer-based “Zurich” using an AM3+ socket. Zurich will herald AMD’s Opteron 3000 Series, targeting those looking for a compact server with cheap consumer-like sockets and will, the company believes, open up a new market of customers pairing cloud computing with the traditional benefits of a local server.

Zurich will arrive in the first half of 2012, while the new AMD Opteron chips are on sale now.

The first Redstone units will be available, albeit in limited volumes, to HP’s business customers next year. HP will also be pushing for a set of Project Moonshot open standards, working with AMD, ARM, Calxeda, Canonical and Red Hat, the target will be so-called “hyperscale” computing environments which balance power and performance with reduced scale and complexity.

Redstone is a major coup for ARM, which bit off two sizable targets in the shape of Intel and AMD when it attempted to challenge the status quo in the enterprise server industry. Until recently, server manufacturers opted for a few high-power chips paired with gobfuls of RAM, a setup that can handle high load but is comparatively wasteful when load reduces.

Chip alternatives like ARM Calxeda take a different approach, using many low-power processors that can scale up to handle larger tasks but use far less power when load is low. HP will also use Intel’s Atom processors in later iterations of Redstone.

The reason the chilly location was chosen is one of cooling and money. The cold climate will allow Facebook to chill its servers using nothing but forced air. The average temperature in the city is 35.6F and hasn’t gone over 86F for more than 24 hours since 1961. The Facebook server farm will be gigantic.

Telegraph.co.uk reports that the massive facility will be built in three buildings covering an entire area the size of 11 football fields. I would assume that since Telegraph is a UK site, they mean soccer fields as far as Americans are concerned. The cold air in the city will help keep costs down, but the costs of running the facility for power alone will be gigantic. The cost for power alone is estimated to be £45M yearly. You have to love the resolve of the person in the photo driving the Mustang in ice and snow.

[via Telegraph]

This new unit will be just under $500 USD and with the ability to begin with any amount of storage, scaling up an option as a person’s usage grows. The DS712+ allows you multiple redundancy options and the two-bay server will support RAID 1 or Synology Hybrid RAID (SHR) for single disk redundancy. Dual NICs will allow for inexpensive MPIO support, SHR will allow users scaling their storage to do so quickly when adding disks in the DX510, and the entire DS712+ runs DiskStation Manager (DSM) software.

DSM software is currently on version 3.2 and supports Mac and Time Machine for OS X Lion. Your DiskStation will include support for joining LDAP domains including simply to act as a lightweight LDAP server. Alongside this is Package Manager which allows you to add packages only when you need them, and users can now browse, install, and run add-ons from inside the DSM. Packages can be installed including VPN server, Syslog Server, and now 3rd party packages as well!

Synology rounds the whole package off with the following features:

Other business-friendly features include:
• Support for ADS domains up to 100,000 users and groups
• ACL support for granular file-level permissions (using local or ADS users and groups)
• iSCSI support with MPIO, MC/S, and SPC-3
• Virtualization support (supports VMware, Citrix Xen, and Hyper-V)

You can find all the places where you can pick up the SD712+ by heading over to Synology.com and tapping the correct region for where you’re living.

However, the outage meant that job applications submitted during a certain time frame were lost. The period that apps were lost in ran from August 7 to about midday on August 9. If you put in an application then you may need to resubmit it. Apparently, the outage happened after a routine maintenance upgrade to the server.

OPM says that no data was compromised or breached because of the crash. Apparently, all of the applications and the agencies looking to fill spots have been made aware of the outage and the people were asked to reapply. This also means that the vacancy period for some positions is being extended. At least this time none of the data was lost and it appears the outage was not the result of hackers.

[via Federaltimes]

PC business revenue for Intel rose by 11 percent year-over-year, indicating that PC sales have yet to be majorly impacted by more mobile developments but have certainly slowed. Their Intel Atom chipset business dropped 15 percent year-over-year, which reflects the decline in netbook sales. This raises concerns in the company’s transition to smartphones and tablets.

However, when asked about Intel’s missing presence in the mobile device market, the company argued that they’re actually already a major mobile device player because they power all of the data centers that deliver content to mobile devices via cloud computing. Indeed, there was a strong demand for Intel’s server chips with their data center group revenue increasing by 15 percent.

Intel CEO Paul Otellini said in a statement that “Strong corporate demand for our most advanced technology, the surge of mobile devices and Internet traffic fueling data center growth, and the rapid rise of computing in emerging markets drove record results.”

[via ZDNet]

The consumer models have a choice of dual-core Intel Core i5 or Intel Core i7 processors, paired AMD Radeon HD 6630M discrete graphics, while the Server model has a quad-core Intel Core i7. $599 gets you the 2.3GHz Core i5 mini with 2GB of RAM and a 500GB HDD, while up to 750GB of hard-drive and 8GB of RAM are options. $799 gets you the 2.5GHz dual-core Core i5, with 4GB of RAM and a 500GB HDD; a 2.7GHz Core i7, 8GB of memory, and either a 750GB hard-drive or 256GB SSD are options.

Finally, the 2.0GHz quad-core Core i7 Mac mini with Lion Server model has 4GB of RAM, twin 500GB hard-drives and starts at $999. Options include up to 8GB of RAM, two 750GB hard-drive or two 256GB SSDs.

Ports and connectivity include Thunderbolt and WiFi. The new Apple Mac mini ships from today.