AMD to sell ARM-based server chips in 2014

Advanced Micro Devices has announced it will sell ARM-based server processors in 2014, ending its exclusive commitment to the x86 architecture and adding a new dimension to its decades-old battle with Intel.

AMD will license a 64-bit processor design from ARM and combine it with the Freedom Fabric interconnect technology it acquired when it bought SeaMicro earlier this year, AMD said Monday.

The result will be a new line of system-on-chip Opteron processors that AMD said will be ideal for the type of massive, web-scale workloads running in giant data centers like those operated by Facebook and Amazon.

AMD Rory Read announces his company’s
ARM move Monday.

AMD CEO Rory Read called the announcement a “seminal moment” and compared it to AMD’s introduction of the first 64-bit x86 processors in 2003. AMD beat Intel to the punch with that move, and it hopes to gain a similar advantage by embracing ARM.

It’s not clear yet if ARM-based CPUs will be successful in servers, but one industry analyst said the move by AMD will help. “I really think this raises ARM’s server credibility, and the credibility of microservers as a segment,” said Patrick Moorhead, president of Moor Insights and Strategy.

Server chips based on the x86 architecture will continue to be the mainstay of AMD’s server business, Read said, but he thinks the ARM-based chips will open up new markets for the company. And while AMD is focused initially on servers, he didn’t rule out the possibility that it will eventually make ARM processors for client devices such as tablets as well.

AMD hopes to sell the new server chips to vendors such as Dell and Hewlett-Packard, and will also sell them in its own servers under the SeaMicro brand. Today those systems are based on x86 processors.

AMD was joined at the event by representatives from Red Hat, Dell, Facebook and (by video) Amazon, a sign of the interest ARM-based server chips are generating.

The timing of Monday’s announcement was a bit awkward, since ARM has yet to unveil the 64-bit processor design that AMD plans to license. It’s likely to be a design code-named Atlas that ARM is expected to unveil at its TechCon conference Tuesday, though neither company would confirm that Monday.

The timing was also bad because hurricane Sandy prevented ARM CEO Warren East from flying in from the UK in time to attend the event. He appeared in a video that was hastily shot in the back of a taxi at Heathrow airport, endorsing the partnership with AMD.

ARM-based servers make sense for the new computing requirements created by services such as social networks and online gaming, said Lisa Su, an AMD senior vice president and general manager. Those workloads need a processor that can efficiently handle very large volumes of small transactions.

“The data center is being inundated with massive amounts of data and there has to be a way to do it more efficiently in a smaller space with a lower cost point,” she said.

ARM architectures are considered more energy-efficient for some workloads because they were originally designed for mobile phones and consume less power. That has attracted several vendors to the space, including Calxeda, Applied Micro and Marvell, all of whom are developing ARM-based chips for servers.

AMD hopes to distinguish itself with two SeaMicro technologies — a custom chip that integrates many components from a traditional server board onto one chip, allowing for dense server designs; and its Freedom Fabric, which can connect thousands of servers in a cluster with low latency and at relatively low cost.

“The fabric technology is the secret sauce; this is what will make AMD’s server solution different from other vendors,” Su said.

Intel has said it won’t make ARM-based processors, in part because it doesn’t want to pay ARM a royalty on each chip. But it has been working hard to reduce the power consumption of its own server chips and said it is confident of its technology roadmap.

The company is due to release a low-power server chip in the second half of the year code-named Centerton, and will follow that up next year with a part dubbed Avoton.

“We have what is required by customers — low powered CPUs, support for key server features, and software compatibility to allow use of current workloads and not force any migration,” Intel spokesman Radek Walczyk said via email.

That still doesn’t give it an equivalent to AMD’s Freedom Fabric, however.

“Think of the chip as half the battle,” said Moorhead, the industry analyst. “The part of the battle [Intel] hasn’t discussed yet is the fabric that makes hundreds or thousands of these parts talk to each other. That’s the magic that guys like Calxeda and AMD are bringing to the table.”

Source: pcworld.com

Read More

IBM's next-gen chips may swap silicon for carbon nanotubes

IBM has hit a milestone in its quest to come up with a successor to silicon computer chips.

The company said Sunday its research into semiconductors based on carbon nanotubes,or CNTs, has yielded a new method to accurately place them on wafers in large numbers. The technology is viewed as one way to keep shrinking chip sizes once current silicon-based technology hits its limit.

IBM said it has developed a way to place over 10,000 transistors made from CNTs on a single chip, two magnitudes higher than previously possible. While still far below the density of commercial silicon-based chips—current models in desktop computers can have over a billion transistors—the company hailed it as a breakthrough on the path to using the technology in real-world computing.

The company made the announcement to mark the publication of an article detailing the research in the journal Nature Nanotechnology.

Intel's latest processors are built using silicon transistors with 22-nanometer technology, and simpler NAND flash storage chips have been demonstrated using "1X" technology somewhere below that, but modern manufacturing is nearing its physical limits. Intel has predicted it will produce chips using sizes in the single digits within the next decade.

Guided by Moore's Law

An IBM scientist shows different solutions
containing carbon nanotubes.

The march toward ever-smaller transistors has produced chips that use less power and can run faster, but can also be made at lower cost, as more can be crammed onto a single wafer. The increasing number of transistors on a given amount of silicon was famously predicted by Gordon Moore, co-founder of Intel, who predicted they would double steadily over time.

Carbon nanotubes, tube-shaped carbon molecules, can also be used as transistors in circuits, and at dimensions of less than 10 nanometers. They are smaller and can potentially 

carry higher currents than silicon, but are difficult to manipulate at large densities.

Unlike traditional chips, in which silicon transistors are etched into circuit patterns, making chips using CNTs involves placing them onto a wafer with high accuracy. Semiconducting CNTs also come mixed with metallic CNTs that can produce faulty circuits, and must be separated before they are used.

IBM said its latest method solves both issues. The company's researchers mix CNTs into liquid solutions that is then used to soak specially prepared substrates, with chemical "trenches" to which the CNTs bond in the correct alignment needed for electrical circuits. The method also eliminates the non-conducting metallic CNTs.

The company said the breakthrough will not yet lead to commercial nano-transistors, but is an important step along the way.

Before they can challenge silicon, however, they must also pass an often-overlooked part of Moore's law—affordability. His law applies to "complexity for minimum component costs," or what consumers are likely to see in the market.

Source: pcworld.com

Read More

PCs with Intel's Haswell chips coming in first half of next year

Haswell will double battery life and performance on ultrabooks compared to Intel's current Ivy Bridge processors

Laptops and desktops with Intel's next-generation Core processor, code-named Haswell, will be available in the first half of next year, Intel CEO Paul Otellini said during a financial conference call on Tuesday.

The Haswell chip will succeed current Core processors code-named Ivy Bridge, which became widely available in April. Intel has said that Haswell will deliver twice the performance of Ivy Bridge, and in some cases will double the battery life of ultrabooks, which are a new category of thin and light laptops with battery life of roughly six to eight hours.

Intel shed some light on Haswell at its Intel Developer Forum trade show in September, saying its power consumption had been cut to the point where the chips could be used in tablets. Haswell chips will draw a minimum of 10 watts of power, while Ivy Bridge's lowest power draw is 17 watts. Intel has splintered future Haswell chips into two families: 10-watt chips for ultrabooks that double as tablets, and 15-watt and 17-watt chips designed for other ultrabooks and laptops.

Haswell will be "qualified for sale" in the first half of 2013, said Stacy Smith, chief financial officer at Intel, during the conference call. Chips go through a qualification process internally and externally, after which Intel can put the chip into production.

The Haswell chip could provide a spark to the ultrabook segment, which has stagnated in a slumping PC market. Worldwide PC shipments dropped between 8 percent and 9 percent during the third quarter, according to research firms IDC and Gartner. They said ultrabook sales were lower than expected due to high prices and soft demand for consumer products.

Many ultrabook models with Ivy Bridge processors are expected to ship in the coming weeks with the launch of Windows 8, which is Microsoft's first touch-centric OS. Otellini said more than 140 Core-based ultrabooks will be in the market, of which 40 will have touch capabilities. A few models -- between five and eight -- will be convertible ultrabooks that can also function as tablets. A majority of the ultrabooks will have prices either at or above US$699, with a few models perhaps priced lower, Otellini said.

The new graphics processor in Haswell will support 4K graphics, allowing for a resolution of 4096 by 3072 pixels. Ultrabooks with Haswell will also include wireless charging, NFC capabilities, voice interaction and more security features.

Otellini said Intel can't tell how the segment will perform in the coming quarter. A number of factors needed to be considered including Microsoft's Windows 8 and the launch of new ultrabooks, he said. Intel reported a profit and revenue decline in the third fiscal quarter of 2012.

"We saw a softening in the consumer segments" in the third fiscal quarter, Otellini said. "The surprise there was China, which was strong, [but] turned weak on us."

Tablets have changed the way people use computers, and Microsoft is bringing touch to mainstream PCs for the first time with Windows 8, Otellini said. PCs with Windows 8 are expected to ship later this month, and it's hard to predict what the response will be until people go out and play with the devices and the OS, Otellini said.

"I see the computing market in a period of transition," with an opportunity for breakthroughs in research and creativity, Otellini said. New usage models for laptops are emerging with detachable touchscreens, voice recognition and other features, and Intel is trying to tap into those opportunities, Otellini said.

The company has a history of overcoming slumps through research and innovation, Otellini said.

Source: computerworld.com

Read More