IBM claims chip breakthrough using carbon nanotubes

IBM has reported that they’re making great strides on developing a new technology that will continue to make chips smaller, while also making them continually faster at the same time. Using carbon nanotubes, IBM scientists have been able to build hybrid chips with more than 10,000 working transistors.

It’s said that the point in time when technology will reach a plateau as far as getting smaller and faster will come at some point, meaning that Moore’s Law won’t last forever. Even though Moore’s Law has lasted almost a half-century so far, it’s said to only be around until around 2020, give or take a few years.

If you’re not familiar with Moore’s Law, it’s basically an observation of sorts where the number of transistors that can fit onto an integrated circuit doubles roughly every two years. The law is named after Intel co-founder Gordon Moore and was coined by computer scientist and former Caltech professor Carver Mead.

IBM’s carbon nanotube discovery is huge, especially considering that chip makers have not yet found ways to improve chips beyond the next two or three generations. Not only will carbon nanotubes allow chip makers to build smaller transistors, but they’ll be able to increase the speed at which those transistors can be turned on and off. However, it’s not said when the new technology will be ready, but it most likely won’t be for a few more years at least, since it’s still in its early development stages.

[via New York Times]

Image via Flickr


IBM claims chip breakthrough using carbon nanotubes is written by Craig Lloyd & originally posted on SlashGear.
© 2005 – 2012, SlashGear. All right reserved.


AMD to develop 64-bit ARM processors for servers

AMD has unveiled its plans to produce 64-bit ARM-based processors for servers, stating that it will be the first company providing x86 and 64-bit ARM solutions. This move aims to satisfy the ever-increasing demand for more energy-efficient 64-bit server processors. The first 64-bit ARM-based processor will be an AMD Opteron, and is slated for production in 2014.

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]


AMD to develop 64-bit ARM processors for servers is written by Brittany Hillen & originally posted on SlashGear.
© 2005 – 2012, SlashGear. All right reserved.


AMD promises 64-bit ARM-based Opteron server CPUs coming in 2014

AMD promises 64-bit ARM-based Opteron server CPUs coming in 2014

AMD has long stuck to x86 architecture for its server processors, but its gearing up to add 64-bit ARM-based Opteron CPUs to its arsenal in 2014. Sunnyvale also plans to reap the fruits of its SeaMicro acquisition by employing the company’s “fabric” tech to link its ARM-based processors in clusters for maximizing efficiency. Where might these new processors come in handy? AMD thinks they’ll fit nicely into clouds and “mega data centers” thanks to their power efficiency, but it’ll let its x86-based hardware do the heavier lifting such as video encoding and rendering. Other details on the CPUs are scarce, but we suspect that’ll change as 2014 approaches.

Continue reading AMD promises 64-bit ARM-based Opteron server CPUs coming in 2014

Filed under: ,

AMD promises 64-bit ARM-based Opteron server CPUs coming in 2014 originally appeared on Engadget on Mon, 29 Oct 2012 19:04:00 EDT. Please see our terms for use of feeds.

Permalink   |  sourceAnandTech  | Email this | Comments

AMD unveils 4GHz+ FX Series Vishera processors

AMD has revealed the latest in its FX Series of processors: the Piledriver Vishera chips. Aimed at computer users with demanding hardware needs, the new series comes in 4, 6, and 8-core variants with a “default” speed of 4GHz. The chips feature a variety of enhancements and improvements over previous offerings.

The new line is composed of four processors: the 4.2GHz FX-8350 (8-core, $195), the 4.0GHz FX-8320 (8-core, $169), the 4.1GHz FX-6300 (6-core, $132), and the 4.0GHz FX-4300 (4-core, $122). The FX-4300 has 4MB of L3 cache, while the FX-6300 has 1MB of L2 cache per core with a total of 8MB of L3 cache. The two higher-end processors, the FX-8320 and FX-8350, both have 8MB of L3 cache. This series uses the Piledriver architecture, which is an improved version of the previously released Bulldozer.

The Piledriver cores were first introduced back in May in the A Series Trinity chips. Improvements include improved scheduling, larger L1 TLB, larger load queue, page translation reload optimization, and more. Said Adam Kozak, “We’re seeing anywhere from seven, all the way up to 15% [in improvement improvement]. Half of that comes from [instructions per cycle] improvements, and the other half or so comes from frequency improvements.” According to AMD, you need a 9-series motherboard to run the FX Vishera chips.

During a pre-release presentation, these four AMD processors were compared to Intel’s Core i5-3570K, i5-3450, i5-2300, and i3-2120 processors. In each case, the AMD processor was priced lower and offered better specs, according to the presentation slide. All FX Series processors are unlocked and can be overclocked.

[via Tech Spot]


AMD unveils 4GHz+ FX Series Vishera processors is written by Brittany Hillen & originally posted on SlashGear.
© 2005 – 2012, SlashGear. All right reserved.


Calxeda announces plans to use 64-bit ARM chips in 2014

Calxeda, which claims to be the first company that implemented ARM processors in servers, has announced its plans to use 64-bit ARM chips in its servers by 2014. The server start-up company detailed its roadmap, which outlines plans for two new generations of hardware built on its own version of 64-bit and 32-bit ARM processors.

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.


Calxeda announces plans to use 64-bit ARM chips in 2014 is written by Craig Lloyd & originally posted on SlashGear.
© 2005 – 2012, SlashGear. All right reserved.


Intel roadmap reveals 10-core Xeon E5-2600 V2 Ivy Bridge CPU

DNP Intel roadmap outs ten core Xeon E52600 V2 ten core Ivy Bridge CPU

Intel may have recently spilled its Q3 guts for 2012, but we highly doubt that the chip maker planned on outing its forthcoming projects for next year. An alleged internal slide makes the claim that the silicon giant plans to introduce a 10-core Xeon E5-2600 V2 Ivy Bridge-EP CPU in the third quarter of 2013. Compatible with Socket R LGA 2011 motherboards, this brute will max out at 20 threads through HyperThreading. Packing 30MB of L3 cache, this unannounced Ivy Bridge supports up to 1866MHz of DDR3 system RAM. If these specifications have whet your appetite, the Xeon E5-2600 V2 is only the tip of the iceberg — Chipzilla is said to also have a 12-core processor in the pipeline as well.

Filed under: ,

Intel roadmap reveals 10-core Xeon E5-2600 V2 Ivy Bridge CPU originally appeared on Engadget on Wed, 17 Oct 2012 21:04:00 EDT. Please see our terms for use of feeds.

Permalink PC Perspective  |  sourceComputerBase (translated)  | Email this | Comments

Amazon in Talks to Buy Texas Instruments’ Mobile Chip Division [Rumors]

Amazon is apparently in “advanced negotiations” with Texas Instruments, in an attempt to buy its smartphone and tablet oriented OMAP chip division. More »

Apple reportedly moving chip production away from Samsung

Apple and Samsung, as many of you already know, have not had the most civil of professional relationships. Samsung has long been a provider of hardware for Apple products, but with the recent high-profile patent suit between the two, it seems that relationship has soured a bit. New reports from CNET and Taiwan Economic News (via MacRumors) are claiming that Apple is looking to move away from Samsung when it comes to the chips for its products, instead striking a deal Taiwan Semiconductor Manufacturing Co. (TSMC) and having that company produce quad-core processors for future devices.


However, this shift in manufacturing partners may not necessarily be because of the bad blood between Apple and Samsung. Apple is interested in TSMC’s 20 nanometer processes – currently, Samsung develops the A-series chips inside the iPhone and the iPad with a 32 nanometer process, and moving to TSMC’s 20 nanometer process would promise chips that are more powerful and efficient. Unfortunately for Samsung, TSMC would become the sole provider of Apple’s chips, which means that Samsung would lose a pretty big customer once Apple begins looking elsewhere.

Analysts confirmed the change in partners to CNET. “Apple is working with TSMC at 20 nanometers,” said Piper Jaffray chip analyst Gus Richard. Another unnamed source did blame the shift on Apple’s soured relationship with Samsung, saying that the iPhone manufacturer is merely looking to fulfill contractual obligations with Samsung before looking elsewhere. Because of that, we’re likely to see some Apple products using Samsung-made chips, while others begin using TSMC chips as Apple slowly switches from one to the other.

While netting Apple as a customer would be a big win for TSMC, this isn’t good news for TSMC’s other customers. Apple, obviously, buys an incredible amount of processors due to the absurd popularity of its devices, so putting that kind of demand on TSMC means that less attention will be paid to the company’s other clients. Apple could look to test out TSMC later this year or early 2013, with production of these new chips ramping up late next year for use in products that will be releasing in 2014. Stay tuned, because we’ll undoubtedly hear more about this change soon.


Apple reportedly moving chip production away from Samsung is written by Eric Abent & originally posted on SlashGear.
© 2005 – 2012, SlashGear. All right reserved.


These Brand New AMD A-Series Processors Could Power Your Next Desktop [Guts]

AMD just announced the processor that could power your next desktop: the second generation A-Series processor. It has more cores for more power, integrated graphics with the AMD Radeon HD 7000 for graphics and puts a high priority on power efficiency. More »

Qualcomm reveals quad-core Snapdragon S4 Play processors, ramps up entry smartphone speeds

Qualcomm reveals quadcore Snapdragon S4 Play processors, ramps up entry smartphone speeds

Don’t think Qualcomm is limiting its quad-core processors to superstar phones. The Snapdragon S4 Play line is growing to include the MSM8225Q and MSM8625Q, parallels to the existing two Play chips that bring four cores to entry-level devices. Besides the speed improvements that you’d expect from all that extra parallelism, the Q variants support the extra bandwidth of low-power DDR2 (LPDDR2) memory and can handle both 720p displays and movie-making. Neither is quite an all-encompassing solution, although the two will cover the bases for much of the starter demographic: while local wireless such as Bluetooth, FM radio and WiFi have to remain separate from the main processor, the two newcomers manage to pack either single-mode UMTS 3G (in the 8225Q) or dual-mode CDMA and UMTS (in the 8625Q) for their cellular fix. Along with the already promised, China-focused S4 Plus MSM8930, test samples of the faster S4 Play editions will be ready before the end of the year, with shipping phones on the way in early 2013 — just in time to go head-to-head with a similar push by MediaTek to make quad-core the norm for a much larger slice of the population.

Continue reading Qualcomm reveals quad-core Snapdragon S4 Play processors, ramps up entry smartphone speeds

Filed under: ,

Qualcomm reveals quad-core Snapdragon S4 Play processors, ramps up entry smartphone speeds originally appeared on Engadget on Wed, 26 Sep 2012 23:14:00 EDT. Please see our terms for use of feeds.

Permalink   |   | Email this | Comments