SI
SI
discoversearch

 Technology Stocks | New Technology


Previous 10 | Next 10 
From: FUBHO1/31/2012 6:09:30 PM
   of 215
 
A mazon S3 Reports Staggering Growth in 2011

By Marshall Kirkpatrick posted by Glenn Peterson
ReadWrite Cloud
January 30, 2012 9:00 PM

Amazon Web Services just reported jaw-dropping growth in the number of objects stored in Amazon S3 year over year.

"As of the end of 2011, there are 762 billion (762,000,000,000) objects in Amazon S3. We process over 500,000 requests per second for these objects at peak times," AWS Evangelist Jeff Bar wrote on the company's blog tonight. The company reported 262 billion objects in storage in Q4 of 2010. "This represents year-over-year growth of 192%; S3 grew faster last year than it did in any year since it launched in 2006." Independent analysts say this is indicative of the growth of the cloud in general and of Amazon's striking dominance of the market.



"Stunning, isn't it?" Randy Bias, co-founder of Cloudscaling said to me about the news by email. "From 150% to almost 200% growth. That's crazy. 500,000 requests per second at peak. Blows my mind." Bias says these are the big take-aways.

- "S3 growth is accelerating, not just increasing. If other AWS services are accelerating similarly then we will see a major shift this year in AWS usage and likely revenue reporting in SEC filings.

- "This is the largest storage system in the world bar none; there isn't anything like it anywhere else that I'm aware of unless it's some secret government/NSA vault.

- "Check my math, but at 1Kbyte average per object, that would be 780PB of disk storage:

- 762,000,000,000 * 1024 (traditional KB)

- 780288000000000 / 1000 (KB for disk) / 1000 (MB for disk) / 1000 (GB for disk) / 1000 (PB for disk) [ disk capacity is in even 1,000 increments, not multiples of 2 ]

- That's 780PB, but unclear if that's replicated or unreplicated; probably replicated, which means 260PB of data with 3x replication.

- Average of 1Kbyte is probably too low.

- At 100TB per storage system that is 7,222 storage *servers*, each with 36 spindles at 3TB each; that might not be their configuration, but even if it's 2 or 3 times as dense, that is a *lot* of storage servers.

- At those numbers, it's a 26M/month business and a 300M/year run rate, which means it's still roughly 30% of AWS revenue with EC2 being most of the rest.

"I don't understand how people can't see this kind of thing and just have their jaw hit the floor. People are paying for this. At this rate they will have 2 TRILLION objects in another year and it will be a $600M/year business."

What's behind such numbers? Widespread technology change.

"What we are seeing is the geometric explosion of cloud growth from multiple points," Constellation Research analyst Ray Wang told ReadWriteWeb.

"First, broad based adoption driven by consumerization of IT. Second, the shift from transaction to engagement - we have social, mobile, analytical, and other unstructured data. Third, true elasticity has come to fruition as the promise of the cloud gets delivered. People are taking to the cloud because the tools are easy to use and they don't have time or money to provision expensive servers. Instead they are using elasticity, which was the original premise of AWS. We could see it happening last year but this leap in growth is tremendous."


Dave Linthicum, CTO and Founder of Blue Mountain Labs, says Amazon's dominance is clear. "The rapid growth of AWS S3 is pretty much in-line with what I'm seeing in enterprises adopting cloud computing. The reality is that they are the 800 pound gorilla, and continue to gain weight. Unless they do something stupid, they are the storage provider to beat."

Ray Wang concurs. "There are only a few companies in the world who can compete with Amazon," he told me by IM tonight.

"It has established itself as one of the leading contenders. The barriers of entry are high. Very few folks can afford to build the data centers, the software infrastructure, and momentum to be profitable. Amazon is in the same league as Google, Microsoft, IBM, etc. The only other folks that could do it if they woke up are the telco's - but we've all been telling them that for years. They haven't paid attention."


Amazon's Barr explains the growth thusly. "Although we definitely made it easier for you to delete objects using Multi-Object Deletion and Object Expiration, we also gave you plenty of ways to upload new objects using Multipart upload, AWS Direct Connect, and AWS Import/Export," he wrote in his blog post. He concluded by noting that running a system so complex is hard work and pointed to open jobs at AWS.

readwriteweb.com 

Share Recommend | Keep | Reply | Mark as Last Read

From: FUBHO2/17/2012 5:16:52 PM
   of 215
 
Oxford Nanopore unveils mini-DNA reader

By Clive Cookson, Science Editor

February 17, 2012 5:09 pm

A young British company has made a powerful entry into one of the fastest-moving and most competitive fields of technology – gene sequencing for science and medicine.



Oxford Nanopore, set up in 2005, unveiled on Friday the world’s first miniature DNA sequencer, small enough to fit in the hand.



The company says its technology reads the four biochemical letters of DNA more quickly and less expensively than the established companies in the field.

The corporate leaders in DNA sequencing are two US companies: Illumina, for which Roche of Switzerland launched a $5.7bn hostile bid last month, and Life Technologies, which made a splash with its recent announcement of a machine that could read a whole human genome – 3bn DNA letters – for just $1,000 in less than a day.

What makes Oxford Nanopore’s “strand sequencing” so effective is that it reads the chemical letters on the DNA directly, one by one, as the molecule ratchets through a microscopic nanopore – a round protein structure with a hole in the middle. Each letter is recognised by its distinct electrical signal.

One advantage of the technique is that it can read much longer strands of DNA than other sequencing methods. Another, said Clive Brown, chief technologist of Oxford Nanopore, is that it requires less sample preparation – making it possible for a doctor to read a patient’s DNA directly from a blood sample in the surgery.

Oxford Nanopore said its MINION sequencer, the size of a USB memory stick, would be available commercially this year at a price below $900.

Continues...

ft.com 

Share Recommend | Keep | Reply | Mark as Last Read

From: Glenn Petersen3/3/2012 4:45:27 AM
   of 215
 
For iRobot, the Future Is Getting Closer

By CHRISTOPHER DREW
Neew York Times
Published: March 2, 2012


iRobot
The iRobot 310 SUGV can gather information in dangerous conditions and lift up to 15 pounds
__________

BEDFORD, Mass. — Ever since Rosey the Robot took care of “The Jetsons” in the early 1960s, the promise of robots making everyday life easier has been a bit of a tease.

Rosey, a metallic maid with a frilly apron, “kind of set expectations that robots were the future,” said Colin M. Angle, the chief executive of the iRobot Corporation. “Then, 50 years passed.”

Now Mr. Angle’s company is trying to do Rosey one better — with Ava, a 5-foot-4 assistant with an iPad or an Android tablet for a brain and Xbox motion sensors to help her get around. But no apron, so far.



With Ava, left, iRobot is trying to do Rosey the Robot of "The Jetsons" one better. Ava will have an iPad or Android tablet for a brain and Xbox motion sensors to help her get around.
____________________

Over the last decade, iRobot, based outside Boston, has emerged as one of the nation’s top robot makers. It has sold millions of disc-shaped Roomba vacuum cleaners, and its bomb disposal robots have protected soldiers in Iraq and Afghanistan. Now, with Ava, it is using video and computing advances to create robots that can do office work remotely and perhaps one day handle more of the household chores.

In late January, iRobot expanded a partnership with InTouch Health, a small company that enables doctors at computer screens to treat stroke victims and other patients from afar. And this week, Texas Instruments said it would supply iRobot with powerful new processors that could help the robots be more interactive and gradually lower their cost.

“We have a firm belief that the robotics market is on the cusp of exploding,” said Remi El-Ouazzane, vice president and general manager of the Texas Instruments unit that makes the processors.

Mr. Angle’s hopes for broadening the industry’s appeal are shared by other robot companies, which have struggled to expand beyond industrial and military uses, toys and other niche products.

Programming robots to mimic human behavior remains difficult. But the ability to use the tablets as simple touch-screen controllers is attracting more software developers, who are envisioning applications that could enhance videoconferencing, provide mobile security guards and sales clerks and help the elderly live longer in their homes.

And with their own innovations now at the center of the effort, the technology giants — Apple, Google, Microsoft and the semiconductor companies — are also pushing things along.

Mr. Angle, 44, who has been at the forefront of robotics since he was a student at M.I.T., said Ava “is one of the things in our pipeline that I am personally most excited about.” But he cautioned that the robot was still a prototype and would not report for any actual work duties before next year.

Mr. Angle estimates that the early versions of Ava will cost in the tens of thousands of dollars, high enough that the company is focusing first on medical applications with InTouch Health, based in Santa Barbara, Calif.

InTouch already has robots with video hookups in many smaller hospitals, and they have saved lives in emergencies when specialists could not get there in person. But the doctors have to drive and manipulate the robots with joysticks to see the patients.

Mr. Angle said that a tap on Ava’s tablet screen could dispatch it to the right room and free doctors from the more mundane controls. Its mapping system, based partly on Microsoft’s 3-D motion sensor for the Xbox, could enable the robot to hustle to the patient’s bedside without slamming into obstacles.

As time goes on, Mr. Angle says he thinks that businessmen could use the robots as proxies at meetings, speaking and watching wirelessly through Ava’s headgear and even guiding her into the hall for private chats. And if the sticker price eventually gets down to consumer levels, as he thinks it will, Ava could, with arms added, dispense pills to baby boomers or even fetch them cocktails.

Still, given how long other robotic breakthroughs have taken, Wall Street is not sure what to make of all this yet.

As sales of its vacuums and military robots grew, iRobot’s earnings shot up to $40 million last year from $756,000 in 2008, and its stock surged to $38 a share from $7. But with pressure mounting for budget cuts at the Pentagon, Mr. Angle told analysts last month that the company’s military sales could drop by as much as 20 percent this year, and the stock quickly tumbled to $25 to $26 a share.

The company had laid off 55 of the 657 employees it had last fall in anticipation of a slowdown in military sales in the United States, and the head of that division departed last month amid concerns that iRobot had not picked up enough military sales to foreign governments.

Frank Tobe, an independent analyst who publishes the Robot Report online, said that until Ava was equipped to pick up and handle objects, the robot would have limited uses. But he said the partnership with InTouch gave iRobot a much-needed toehold in health care. iRobot plans to invest $6 million in InTouch, and Mr. Tobe said by combining their technologies, the companies could produce devices at a much lower cost and attract more business.

IRobot also faces growing competition from robotics companies in Asia and Europe, many subsidized by governments that believe the innovations will help push their economies forward. But analysts say iRobot has a number of crucial patents. And the company has a strong track record in finding practical uses for robots and getting them to market.

Mr. Angle’s first robot, built in the late 1980s with Rodney Brooks, an M.I.T. professor, was Genghis, a buglike creature that ended up in the Smithsonian. Powered by microprocessors with only 156 bytes of memory, it could walk on six legs. It also showed that robots could be programmed to react to just a few basic rules.

That project piqued Mr. Angle’s interest in building simple, practical robots. He, Dr. Brooks and another M.I.T. graduate, Helen Greiner, started iRobot in 1991, he said, “to make robots that would touch people’s lives on a daily basis.”

But that goal proved harder than they expected, and a decade of trial and error followed. Standing by a display here at the company’s headquarters, Mr. Angle pointed to some of its early efforts, including a robotic doll for Hasbro called My Real Baby and little wooly blue and orange creatures that could scurry and hide.

But, he said, “from the very first moments of iRobot, whenever I would introduce myself to someone on an airplane or wherever, the response nearly 100 percent of the time was not ‘How are you?’ but ‘When are you going to clean my floors?’ They wanted Rosey from ‘The Jetsons.’ ”

“So very, very early on, we knew cleaning was a great application, if only we could figure out how to do it,” he added.

But it was not until 2002 that everything came together, with the introduction of the Roomba vacuum and an urgent military demand for robots that could check out dangerous caves in Afghanistan. Those 50- to 60-pound robots, called Packbots, also turned out to be critical in Iraq in disarming roadside bombs and acting as sentries at checkpoints.

Since then, sales of new versions of the Roomba, which cost $350 to $600 each, have taken off, especially overseas. The company has started selling robots for cleaning bathroom floors, called Scooba, for $280 to $500. It has also developed lightweight robots with video cameras that soldiers can toss into windows before storming a building. They include a 30-pound model and a tiny new five-pounder, called FirstLook, now being tested in Afghanistan. And even if their orders slow, top Pentagon officials remain committed to robots to save money and soldiers’ lives.

The company’s goal, Mr. Angle said, continues to be building robots that can operate more autonomously or provide “remote presence” — tech-speak for enabling people to be in two places at one time.

(Mr. Angle knows something of that language. After he appeared in 2008 as an M.I.T. professor in a film with Kevin Spacey called “21,” the director said he had gotten just what he wanted from Mr. Angle. “You know, you just can’t coach geek.”)

Mr. Angle said he, too, was looking forward to the day when robots like Ava would have arms and even keener sight.

“I like the idea that if you have a party, the robot can recognize faces, take drink orders, go back to the kitchen, load it up and then go back and find those people and deliver the drinks,” he said. “I think that would be awesome.”

nytimes.com 

Share Recommend | Keep | Reply | Mark as Last Read


From: FUBHO3/19/2012 2:25:10 PM
   of 215
 
Chip Makers Mull Plans to Insert DSA at 14nm

By Mark LaPedus and David Lammers

semimd.com 

Faced with the dreaded multi-patterning era and delays with extreme ultraviolet (EUV) lithography, chip makers are taking a harder look at a technology that could save the day for the industry: directed self-assembly (DSA). In fact, several IC vendors are mulling plans to implement DSA at the 14nm node or so — an insertion point that is sooner than previously thought, according to one DSA materials supplier.

For some time, there has been a consensus that the IC industry would use 193nm lithography and multi-patterning at 14nm. Then, at 10nm, the IC industry could go with several options: EUV, maskless, multi-patterning or nano-imprint.

Used in conjunction with today’s 193nm lithography scanners, DSA has also been seen as a possible candidate for IC production at the 10nm node. DSA — an alternative patterning technology that makes use of block copolymers — could also turn the IC industry upside down. It could extend 193nm wavelength lithography beyond 10nm, potentially eliminate expensive multi-patterning steps, and possibly push out EUV.

In one example of its enormous potential, DSA, along with 193nm immersion lithography, has demonstrated the ability to print lines and spaces down to 12.5nm — without the need for multi-patterning, said Christopher Bencher, a member of the technical staff at Applied Materials Inc., in a recent interview.

“Some predictions play it safe and target the 10nm node” for DSA, said Ralph Dammel, chief technology officer (CTO) for AZ Electronic Materials, a supplier of materials for DSA and other applications. “However, we see a lot of customer interest for insertion already at 14nm and even higher nodes,” Dammel said. “Teams have been assigned at major customers, not just to study the potential of DSA, but to ready it for near term insertion. My best guess is that we will see some penetration of DSA by 2013/14, and that at the 10nm node, it will become pervasive.”

The early DSA programs among some undisclosed chip makers are expected to kick off this August, which is geared for the 14nm node in the 2015 time frame, he said.

But current block copolymers based on today’s poly(MMA-co-styrene) technology could hit the wall somewhere between 14nm to 10nm. As a result, there is a wave of research to find new copolymers that could extend the technology to 10nm and beyond.

AZ Electronic, Dow, JSR, SEH, TOK and others are racing each other to develop next-generation DSA materials. AZ Electronic, which claims to be the leader in DSA materials, is expected to disclose its new results with the technology at the Semicon China trade show in Shanghai from March 20-22.

DSA is capturing the imagination of the IC industry, but experts are quick to point out that the technology faces some challenges and there is a major debate when it will be ready for prime time. Moshe Preil, manager of emerging lithography and tools at GlobalFoundries Inc., said the industry has taken a “more serious and harder look” at DSA since the SPIE Advanced Lithography event in February. At SPIE, there were several troubling disclosures, namely that EUV and the associated power sources are still far behind the curve.

Preil, who runs the DSA program at GlobalFoundries, stopped short of saying that DSA will be inserted at the 14nm node. The insertion point largely depends upon the progress with the technology, he said.

During a panel session at this week’s Common Platform Technology Forum 2012 in Santa Clara, Calif., T.C. Chen, IBM Fellow and vice president of science and technology at IBM Research, said: “DSA is coming in pretty soon for critical levels. Triple patterning is not really economically feasible.”

During another session at the same event, Lars Liebmann, distinguished engineer at IBM, had a different view, saying that adding “one more mask layer to a complex mask set” for triple patterning is not going to stop companies from going that route if they need it to pattern the critical layers.

Liebmann said DSA could be used at the 10nm node to “make the gratings, which could then be cut with an e-beam. It is not an option for 14nm; it just won’t be ready in time.”

DSA — Next big thing?

Still, GlobalFoundries, Hynix, IBM, Intel, Micron, Samsung, TSMC and others have begun to take DSA more seriously – and for good reason. EUV lithography is late. Maskless and nano-imprint lithography are also not ready.

So, leading-edge chip makers must continue to use today’s 193nm immersion lithography tools for advanced chip production, but they are also forced to implement expensive multi-patterning steps. “Multi-patterning will be here for some time — at least for two more nodes,” said Michael White, director of product marketing for Calibre Physical Verification at Mentor Graphics Corp.

For 20nm, the industry must embrace double-pattering in one form or another. “Triple-patterning will not happen at 20nm,” White said at the Common Platform event. “It’s one of the options for customers at 14nm.”

EUV is largely required because it brings the industry back to single-exposure technology. But if EUV misses its target window at 14nm and emerges at 10nm, White sketched out a troubling scenario: At 10nm or beyond, the industry may end up using EUV and double-patterning simultaneously.

As a result, many hope DSA can save the day. In 2007, DSA landed on the International Technology Roadmap for Semiconductors (ITRS) roadmap as a potential solution for lithography at the 10nm node.


IBM's DSA process flow (Source: AZ Electronic)


DSA is not a next-generation lithography (NGL) tool, but it is actually a “complementary” technology. DSA is an alternative patterning technology that enables frequency multiplication through the use of block copolymers. When used in conjunction with an appropriate pre-pattern that directs the orientation for patterning, DSA can reduce the pitch of the final printed structure.

At SPIE, Applied’s Bencher said defectivity, registration and other issues remain some of the key challenges to move DSA into production. DSA is ideal for dense contacts, Fin patterning and other applications, he said.

AZ Electronic, IBM, the University of Wisconsin and others have separately developed rival DSA “process flows” for chip production. IMEC has recently announced the implementation of the world’s first 300mm fab-compatible DSA process line.

Saving the day

Mukesh Khare, director of semiconductor technology research at IBM Research, said some people refer to DSA as “pitch in a bottle.” He himself referred to DSA as “polymer self-assembly” and said IBM has been working on polymers for quite a long time.

Khare said DSA has already been used to produce 25nm line and spaces with good line-edge roughness, using immersion tools. “Lithography defined directing patterns” are capable of 10nm resolution, he said at a session during the Common Platform event. The work is very encouraging, he added, with “defectivity similar to when we first started to play around with immersion lithography.”


(Source: AZ Electronic)


The polymers include materials with much different molecular weights, which separate at various phases. “We are trying to determine the self-assembly morphology,” he added.

DSA is “like the early days of immersion lithography, when there was a similar feeling of exhilaration,” said AZ Electronic’s Dammel. “I won’t go as far yet as to predict that EUV will share the fate of 157nm lithography. But clearly, DSA is on its way to becoming a mainstream, low cost lithography option.”

In 2010, Luxembourg-based AZ Electronic signed an agreement with IBM to co-develop DSA technology. IBM also has a separate deal with JSR Corp. in the arena.

At Semicon China, AZ Electronic will announce that it can reproduce “synthesize performing block copolymers with half-pitch at 10.5nm to 31nm” (See images below). The company said it has completed the first stage of materials learning and samples are ready for ”in-fab process learning.” The company’s block copolymers supports AZ’s own process flow as well as those from IBM and the University of Wisconsin.


AZ tips its own process flow for DSA (Source: Company)


“By late 2011, we had achieved this target,” Dammel said. “IMEC has received materials for all of these flows for use in their DSA pilot line, and we are currently the only supplier who is able to provide gallon samples to them. These samples are low metal, properly filtered, lithographically tested for DSA performance, and meet all requirements for IC production. We now stand ready to provide these materials in quantities for process learning and production insertion, and I think no one else can say that at present.”

Now, the trick is to move DSA from the lab to the fab at or around 14nm. Then, current polymer technology could hit the wall, fueling a new wave of R&D for next-generation materials. “For 10nm, p(MMA-co-styrene) block polymer is no longer a suitable material,” he said. “Its low chi factor implies that a high MW is needed to obtain phase separation, and since MW is related to domain size, the lowest line/space structures that can reliably be made are (about) 11nm.”

AZ and others are developing higher chi materials for 10nm node and below. “With these materials, it will be easily possible to extend DSA to the 8nm node, using guide structures made by immersion lithography using the trick of putting more than one polymer stripe to the guide structure,” he said.

There are a number of promising leads for the newfangled polymer. For example, the University of Queensland in Australia is developing a new class of diblock copolymers called PS-b-PDLA. With PS-b-PDLA, “8nm node type features (are) possible,” he said.

There are at least two tool options for 8nm: 193nm lithography and DSA and/or EUV and DSA. “We may see a co-existence between EUV and DSA,” he said. “But if EUV up and dies, the world doesn’t end.”


AZ shows images using DSA (Source: Company)

Share Recommend | Keep | Reply | Mark as Last Read | Read Replies (1)

To: FUBHO who wrote (134)3/19/2012 5:37:42 PM
From: Brian Sullivan
   of 215
 
Amazon.com to Acquire Kiva Systems, Inc.

SEATTLE--(BUSINESS WIRE)-- Amazon.com, Inc. ( AMZN) today announced that it has reached an agreement to acquire Kiva Systems, Inc., a leading innovator of material handling technology.

Amazon has long used automation in its fulfillment centers, and Kivas technology is another way to improve productivity by bringing the products directly to employees to pick, pack and stow, said Dave Clark, vice president, global customer fulfillment, Amazon.com. Kiva shares our passion for invention, and we look forward to supporting their continued growth.

For the past ten years, the Kiva team has been focused on creating innovative material handling technologies, said Mick Mountz, CEO and founder of Kiva Systems. Im delighted that Amazon is supporting our growth so that we can provide even more valuable solutions in the coming years.

Following the acquisition, Kiva Systems headquarters will remain in North Reading, Massachusetts.

Under the terms of the agreement, which has been approved by Kivas stockholders, Amazon will acquire all of the outstanding shares of Kiva for approximately $775 million in cash, as adjusted for the assumption of options and other items. Subject to various closing conditions, the acquisition is expected to close in the second quarter of 2012.


4:54 PM
Amazon ( AMZN) -1.1% AH after agreeing to buy privately-held Kiva Systems, which makes robotic technology for warehouse shelf and pallet systems, for $775M in cash.

5:15 PM More on Amazon/Kiva: Kiva has been growing at an 80% clip thanks to an innovative robot/software solution that automates "the process of picking, packing and shipping" products. Founder Mick Mountz boasts Kiva's solution, which has an average price of $5M, can handle 2x-4x as many orders/hour as a traditional approach. One unanswered question is the extent to which Amazon will support Kiva customers such as Staples, Gap, Saks, and Walgreen. AMZN now +0.1% AH. ( PR) [ Tech, M&A] 1 Comment

I was always interested in possibly investing in Kiva, but it was privately held.

There are lots of cool Kiva robotic videos out there of YouTube:

Warehouse Robots at Work:
youtube.com 

Share Recommend | Keep | Reply | Mark as Last Read | Read Replies (1)

To: Brian Sullivan who wrote (135)3/28/2012 6:24:56 PM
From: FUBHO
   of 215
 
Panel ponders many-core ICs tripping 'the singularity'



Peter Clarke
3/28/2012 12:09 PM EDT

SAN JOSE, Calif. – Will the rapidly increasing processing power being enabled by many-core processors cause the advent of machines with super-human intelligence, an event sometimes referred to as the singularity?

That was the question put to a panel of some of the best minds in multicore processor theory and design assembled on Tuesday (March 27) by analyst Jon Peddie at the Multicore DevCon, part of DESIGN West being held here this week. A small but enthusiastic audience was there to listen.

Peddie set up the debate by referencing Vernor Vinge, a science fiction writer, who had predicted that computing power would be equivalent to human processing power by about 2023. One particular aspect of the concept of the singularity is that once machines either in single units or collectively exceed human intelligence there may be an explosion of machine learning advancement that it would not be possible for humans to fathom, by definition, making the singularity a kind of event horizon.

Another extrapolation of computing progress had 2045 as the year in which it might be possible to buy a machine with the processing power of human brain for $2,000 in 2045.

Pradeep Dubey of Intel parallel computing labs illustrated the progress by saying that a petaflops supercomputer can already simulate a cat's brain. A human brain has 20 to 30 times more neurons and 1,000 times more synapses, he said, so the complete simulation of human brain is only a matter of a 5 or 6 years away. "Exaflops could simulate a human brain," he said.


Dubey said there are currently three approaches: simulate the process with a neuron- and synapse-level model; ignore brain architecture and treat the problem as data and statistical problem; or to build hardware that mimics neurons and synapses.

However, the simulation of the brain is not the same as thinking or generating the emotional intelligence we see in human beings, said Ian Oliver director of Codescape development tools at processor licensor Imagination Technologies Group plc. "We probably have the wrong memory model. The human brain is non-deterministic. It operates on the edge of chaos," he said.

Oliver pointed out that the use of genetic algorithms to derive FPGA designs through evolution produced much more brain-like architectures but were not readily usable in the real world and as such computer and human intelligence appeared to be distinct.

Mike Rayfield, vice president of the mobile business unit at Nvidia argued that the number of processor cores is a red herring. But Intel's Dubey countered that more cores is better saying that massive data engines can capture correlations if not causality. He pointed out that machines can already do some things far better than humans, which is the reason they exist. "We can build planes but we can't build birds," he said.


Next: computers already smarter than humans -- at specific things


eetimes.com 

Share Recommend | Keep | Reply | Mark as Last Read


From: FUBHO4/17/2012 12:43:06 PM
   of 215
 

Boron-treated carbon nanotubes soak up oil from water repeatedly April 17, 2012


[+] This carbon nanotube sponge created at Rice University can hold more than 100 times its weight in oil. Oil can be squeezed out or burned off, and the sponge reused. (Credit: Jeff Fitlow/Rice University)


Researchers at Rice University and Penn State University have discovered that adding a dash of boron to carbon while creating nanotubes turns them into solid, spongy, reusable blocks that have an astounding ability to absorb oil spilled in water.

That’s one of a range of potential innovations for the material created in a single step. The team found for the first time that boron puts kinks and elbows into the nanotubes as they grow and promotes the formation of covalent bonds, which give the sponges their robust qualities.

Lead author Daniel Hashim, a graduate student in the Rice lab of materials scientist Pulickel Ajayan, said

The blocks are both superhydrophobic (they hate water, so they float really well) and oleophilic (they love oil). The nanosponges, which are more than 99 percent air, also conduct electricity and can easily be manipulated with magnets.

To demonstrate, lead author Daniel Hashim, a graduate student in the Rice lab of materials scientist Pulickel Ajayan, dropped the sponge into a dish of water with used motor oil floating on top. The sponge soaked it up.

He then put a match to the material, burned off the oil and returned the sponge to the water to absorb more. The robust sponge can be used repeatedly and stands up to abuse; he said a sample remained elastic after about 10,000 compressions in the lab. The sponge can also store the oil for later retrieval, he said.

“These samples can be made pretty large and can be easily scaled up,” said Hashim, holding a half-inch square block of billions of nanotubes. “They’re super-low density, so the available volume is large. That’s why the uptake of oil can be so high.” He said the sponges described in the paper can absorb more than a hundred times their weight in oil.

Ajayan, Rice’s Benjamin M. and Mary Greenwood Anderson Professor in Mechanical Engineering and Materials Science and of chemistry, said multiwalled carbon nanotubes grown on a substrate via chemical vapor deposition usually stand up straight without any real connections to their neighbors.

But the boron-introduced defects induced the nanotubes to bond at the atomic level, which tangled them into a complex network. Nanotube sponges with oil-absorbing potential have been made before, but this is the first time the covalent junctions between nanotubes in such solids have been convincingly demonstrated, he said.

“The interactions happen as they grow, and the material comes out of the furnace as a solid,” Ajayan said. “People have made nanotube solids via post-growth processing but without proper covalent connections. The advantage here is that the material is directly created during growth and comes out as a cross-linked porous network.

“It’s easy for us to make nano building blocks, but getting to the macroscale has been tough,” he said. “The nanotubes have to connect either through some clever way of creating topological defects, or they have to be welded together.”

“Our goal was to find a way to make three-dimensional networks of these carbon nanotubes that would form a macroscale fabric — a spongy block of nanotubes that would be big and thick enough to be used to clean up oil spills and to perform other tasks,” said Humberto Terrones of Oak Ridge National Lab.

“We realized that the trick was adding boron — a chemical element next to carbon on the periodic table — because boron helps to trigger the interconnections of the material. To add the boron, we used very high temperatures and we then ‘knitted’ the substance into the nanotube fabric.”

Oil-spill remediation and environmental cleanup

The researchers have high hopes for the material’s environmental applications. “For oil spills, you would have to make large sheets of these or find a way to weld sheets together (a process Hashim continues to work on),” Ajayan said.

“Oil-spill remediation and environmental cleanup are just the beginning of how useful these new nanotube materials could be,” added. “For example, we could use these materials to make more efficient and lighter batteries. We could use them as scaffolds for bone-tissue regeneration. We even could impregnate the nanotube sponge with polymers to fabricate robust and light composites for the automobile and plane industries.”

Hashim suggested his nanosponges may also work as membranes for filtration.

“I don’t think anybody has created anything like this before,” Ajayan said. “It’s a spectacular nanostructured sponge.”

The paper’s co-authors are Narayanan Narayanan, Myung Gwan Hahm, Joseph Suttle and Robert Vajtai, all of Rice; Jose Romo-Herrera of the University of Vigo, Spain; David Cullen and Bobby Sumpter of Oak Ridge National Laboratory, Oak Ridge, Tenn.; Peter Lezzi and Vincent Meunier of Rensselaer Polytechnic Institute; Doug Kelkhoff of the University of Illinois at Urbana-Champaign; E. Muñoz-Sandoval of the Instituto de Microelectrónica de Madrid; Sabyasachi Ganguli and Ajit Roy of the Air Force Research Laboratory, Dayton, Ohio (on loan from IPICYT); David Smith of Arizona State University; and Humberto Terrones of Oak Ridge National Lab and the Université Catholique de Louvain, Belgium.

The National Science Foundation and the Air Force Office of Scientific Research Project MURI program for the synthesis and characterization of 3-D carbon nanotube solid networks supported the research.

Ref.: Daniel P. Hashim, Covalently bonded three-dimensional carbon nanotube solids via boron induced nanojunctions, Nature Scientific Reports, 2012 [DOI: 10.1038/srep00363] (open access)

Share Recommend | Keep | Reply | Mark as Last Read

From: FUBHO4/24/2012 3:29:25 PM
   of 215
 

Status Quo for ReRAM


• In-depth understanding of resisKve switching mechanism, formaKon energy for
nano-conducKve ?lament: Good progress made
• On-state and o?-state conducKon mechanisms: Progress for on-state, but s*ll a
lot of room for o?-state
• E?ects of electrode and addiKonal element doping: ac*ve electrode vs inert
electrode, but less systema*c work for doping
• Variability of switching characterisKcs: beDer in smaller device area
• Programming power vs data retenKon: subject for ?lament forma*on energy
• E?ort toward scalability: geometry driven vs MLC and/or MLS
• DemonstraKon of cross point memory: selec*on device becoming a focal point
• Product introducKon: Produc*za*on has started with small scale integra*on

ewh.ieee.org 

Share Recommend | Keep | Reply | Mark as Last Read

From: FUBHO5/18/2012 10:32:42 AM
   of 215
 

Samsung researchers propose graphene 'barristor'

eetimes.com 

Peter Clarke5/18/2012 9:01 AM EDT

LONDON – A research team from the Samsung Advanced Institute of Technology (Yongin, Korea) has proposed a novel three-terminal device that could overcome previous problems integrating graphene into circuits. The term barristor comes from running together "variable barrier transistor."

Graphene, an atomic monolayer form of carbon, has attracted much attention because it offers about 200 times higher electron mobility than silicon. However, until recent disclosures about semiconducting graphene monoxide, graphene and its derivatives had only existed as conductors and insulators. The metallic nature of graphene and its high conductivity were accompanied by a lack of hysterisis and no mechanism to switch conduction off.

Converting graphene into a semiconductor can decrease the electron mobility of graphene negating the benefits and leading to skepticism over the feasibility of graphene transistors, Samsung said in a statement.

The team from SAIT, the central R&D arm of Samsung Electronics Co. Ltd., published the paper Graphene Barristor, a Triode Device with a Gate-Controlled Schottky Barrier online in the journal Science on May 17.

The claim is that SAIT has developed a device that can switch off the current in graphene without degrading its mobility. The key is an atomically sharp interface between graphene and hyrogenated silicon. An on/off ratio current modulation of 10^5 is achieved by adjusting the gate voltage to control the graphene-silicon Schottky barrier.

The researchers report the fabrication of complementary p- and n-type graphene barristors as well as inverter and half-adder logic circuits on 150-mm diameter wafers. SAIT said it owns nine major patents related to the structure and the operating method of the graphene barristor.

Share Recommend | Keep | Reply | Mark as Last Read


From: FUBHO6/4/2012 12:50:37 PM
   of 215
 

Neuroscientists reach major milestone in whole-brain circuit mapping projectJune 1, 2012


[+]Projections from a motor cortex AAV injection (credit: CSHL)


Neuroscientists at at Cold Spring Harbor Laboratory (CSHL) reached an important milestone today, publicly releasing the first installment of data from the 500 terabytes so far collected in their pathbreaking project to construct the first whole-brain wiring diagram of a vertebrate brain, that of the mouse.

The data consist of gigapixel images (each close to 1 billion pixels) of whole-brain sections that can be zoomed to show individual neurons and their processes, providing a “virtual microscope.”

The images are integrated with other data sources from the web, and are being made fully accessible to neuroscientists as well as interested members of the general public. The data are being released pre-publication in the spirit of open science initiatives that have become familiar in digital astronomy (e.g., Sloan Digital Sky Survey) but are not yet as widespread in neurobiology.

Each sampled brain is represented in about 500 images, each image showing an optical section through a 20 micron-thick slice of brain tissue. A multi-resolution viewer permits users to journey through each brain from “front” to “back,” and thus enables them to follow the pathways taken through three-dimensional brain space by tracer-labeled neuronal pathways. The tracers were picked to follow neuronal inputs and outputs of given brain regions.

“We’re executing a grid-based “shotgun” strategy for neuronal tract tracing that we first proposed a few years ago, and which I am pleased to note has gained acceptance elsewhere within the neuroscience community,” says Partha P. Mitra, Ph.D., the Crick-Clay Professor of Biomathematics at CSHL and director of the Mouse Brain Architecture (MBA) Project. After the initial June 1 release, project data will be made public continuously on a monthly basis, Mitra says.


[+]Composite image generated with Mouse Brain Architecture project data. Injections of two fluorescently marked (red and green) adeno-associated viral (AAV) tracers indicate neural pathways, superimposed upon a whole-brain image stained to reveal the protective sheathing around myelinated axons. Axonal paths leaving the injection site are seen, including horizontal ones crossing over to the other side of the brain along the Corpus Callosum. (Credit: CSHL)


Project addresses a large gap in knowledge

“Our project seeks to address a remarkable gap in our knowledge of the brain,” Mitra explains. “Our knowledge of how the brain is wired remains piecemeal and partial after a century of intense activity. Francis Crick and Ted Jones emphasized this in an article published in Nature nearly 20 years ago.

Yet to understand how the brain works (or fails to work in neurological or neuropsychiatric disease), it is critical that we understand this wiring diagram more fully.

Further, there remain fundamental questions about brain evolution that cannot be addressed without obtaining such wiring diagrams for the brains of different species.”

The MBA Project, which has received critical funding from the Keck Foundation and from the National Institutes of Health, is distinguished by the approach advocated by Mitra and colleagues in a position paper published in 2009. Mitra there proposed mapping vertebrate brains at what he calls the “mesoscopic” scale, a middle-range amenable to light microscopy, providing far more detail than, for instance, MRI-based methods, and yet considerably less detail than is achievable via electron microscopy (EM).

The latter approach, while useful for mapping synaptic connections between individual neurons, is feasible on a whole-brain basis only for very small brains (e.g. that of the fruitfly) or very small portions of the mouse brain.

The pragmatic approach Mitra advocated and which is realized in this first data release, is to image whole mouse brains in a semi-automated, quality-controlled process using light microscopy and injected neural tracers (both viruses and classically used tracer substances). While the basic methodology has been available for some time, systematically applying it to a grid of locations spanning the entire brain, and digitizing and re-assembling the resulting collection of brains, is a new approach made feasible by the rapidly falling costs of computer storage.

A single mouse brain at light-microscope resolution produces about a terabyte (1 trillion bytes, or 1000 GB) of data; thus, generating and storing the data set currently being gathered would have been prohibitively expensive a decade or so ago.

Assembling the circuit diagram at a mesoscopic scale using ‘shotgun approach’

A key point is that at the mesoscopic scale, the team expects to assemble a picture of connections that are stereotypical – that is, essentially the same in different individuals, and probably genetically determined in a species-specific manner.

By dividing the volume of a hemisphere of the mouse brain into 250 equidistant, predefined grid-points, and administering four different kinds of tracer injections at each grid point — in different animals of the same sex and age — the eight-member team at CSHL assisted by collaborating scientists at Boston University, MIT and the University of California, San Diego seeks to assemble a complete wiring diagram that will be stitched together from the full dataset.

The project in this sense is analogous to the Human Genome Project’s “shotgun” approach, in that its final product – a comprehensive wiring diagram – will be the product of many individually obtained data components, woven together thanks to the power of advanced computing and informatics. Indeed, Mitra says one of the genome project’s early advocates, Dr. James D. Watson (now CSHL Chancellor Emeritus), provided him with motivation and encouragement to pursue the project.

“We will never understand how the brain works until we have the wiring diagram,” Dr. Watson comments today. “Mitra is on the right track and I’m impressed he’s gone from conception to putting out data in a couple of years on a quite modest budget. His approach deserves strong funding support.”

The MBA Project was also inspired by early efforts of the Allen Institute, funded by Microsoft co-founder and philanthropist Paul Allen, which resulted in assembly of a comprehensive map of gene expression across the mouse brain. That effort was the product of standard molecular biology procedures iterated in a quasi-industrialized process. The resulting whole-brain gene-expression map, while a triumph, was not designed to shed light on connections in the brain, which became a point of departure for Mitra.

Since the 2009 publication of Mitra and colleagues’ proposal for meso-scale circuit-mapping projects for whole vertebrate brains, the approach has not only spawned Mitra’s CSHL project, but also other meso-scale circuit-mapping projects for the mouse at the Allen Institute and at UCLA. Each differs in aim and technical detail.

A number of features distinguish the “meso-scale” circuit project at CSHL. The 20-micron spacing between brain “slices” gives the CSHL results a particularly rich sense of three-dimensional depth and detail. The team’s use of four tracers including both classical tracer substances as well as neurotropic viruses (attenuated or disabled viruses that infect nerve cells), provides redundancy and helps control for differing efficacies of the different tracer substances.

The images one sees on the MBA Project website begininng today provide hard data on actual neuronal processes – the “ground truth” of neuroanatomy, in Mitra’s words — and do not rely on inferential methodologies such as functional MRI scans and diffusion tensor imaging to suggest areas in which connections occur. Finally, it is noteworthy that the slides generated by the project are being physically stored, to permit re-examination at a later date, using more refined imaging methods if necessary or as new methods become available.

“Our project is what I’d call a necessary first step in a much larger enterprise, that of understanding both structure and dynamics of the vertebrate, and ultimately, the human brain,” says Mitra. “While facile comparisons with Genome projects should be avoided, the data sets generated by the MBA and similar projects will provide a useful framework — not unlike a reference genome — on which we can ‘hang’ all kinds of neuroscience knowledge, the body of which has always been notably fragmentary.”

The planning stage of the project, including meetings at the Banbury conference center at CSHL as well as initial informatics work, was made possible by an award from the Keck Foundation. Major funding for the Mouse Brain Architecture project comes from a Challenge Grant from the National Institutes of Health (RC1MH088659) and a Transformative Award from the Office of the NIH Director (R01MH087988). Additional sources of funding include internal funding at Cold Spring Harbor Laboratory and the Crick-Clay Professorship.

Ref.: Technical White Paper: Mouse Brain Architecture Project, Cold Spring Harbor Laboratory, Feb. 2012 (open access)

Ref.: Jason W. Bohland et al., A Proposal for a Coordinated Effort for the Determination of Brainwide Neuroanatomical Connectivity in Model Organisms at a Mesoscopic Scale, PLoS Computational Biology, 2009, DOI: 10.1371/journal.pcbi.1000334 (open access)

Share Recommend | Keep | Reply | Mark as Last Read
Previous 10 | Next 10 

Copyright © 1995-2014 Knight Sac Media. All rights reserved.Stock quotes are delayed at least 15 minutes - See Terms of Use.