Al and Ziggy

Researchers at Yale have created the first ever fully functional quantum processor. Harnessing the bizarre qualities of quantum mechanics, the processor can perform simple calculations.

Typical computers use electrons (through transistors) to compute—reading and writing information in bits. Bits have binary states; they’re either “on” or “off,” 1 or 0. Quantum computers use atoms and “qubits,” which have multiple states. Qubits can be 1, 0, 1-0, 0-1, 0+1, or 0 AND 1 simultaneously. Thus a single qubit can store much more information than a bit. Additionally, typical computers read and write numbers and solve problems sequentially. Quantum computers can read and write long strings of numbers all at once, boosting speed tremendously.

The Yale computer is made up of two artificial atoms—billions of aluminum atoms that act as a single atom—in a solid-state system. The processor is extremely unstable, capable of hanging around for only a millisecond before evaporating. Still, it’s a major breakthrough in quantum computing that will lead to more stable and capable computers in the future.

Because of their tremendous computing power and speed, quantum computers have the potential to truly revolutionize computing.

Link to TG Daily article

Two teams of U.S. scientists have pushed Moor’s Law into overdrive, crafting transistors and memory storage material on the nano scale. 

Jeremy Levy and his team at the University of Pittsburgh have created two-nanometer transistors out of lanthanum aluminate and strontium titanate. Levy used an atomic force microscope to etch a miniscule wire between the two insulators, creating the world’s smallest transistor. Even more intriguing, the team was able to use the microscope to “erase” the wire. Using this technique, the team could conceivably reconfigure the transistor to make memory modules. 

Levy says he got the idea for the transistor from an Etch A Sketch, which uses a stylus to scrape aluminum powder off a glass plate.

The new transistor is several times smaller than the smallest silicon transistor (currently 45 nanometers).

Even better, Levy says that atomic force microscopes could be miniaturized down to the size of a wristwatch. Not that you’d ever want to set up a nano transistor factory on your wrist, but still.

Meanwhile, engineers/scientists/all-around-good-guys-and-gals at the University of Massachusetts Amherst and the University of California Berkeley have created a method of reliably making thin-film polymer memory material. Others have tried to make polymer thin-film memory material, but it usually falls apart when spread over large surfaces.

The teams at Amherst and Berkeley used a lattice of sapphire crystals to make a grid to lay the thin-film sheet on. This makes nearly perfect arrays of film that’s 15 times denser than anything that’s ever been made before. And we’re talking dense, about 250 DVDs worth of data on a surface the size of a quarter.

Behold your Young Lady’s Illustrated Primer. Or at least the bits and pieces that could make one.

Link to University of Pittsburgh press release

Link to Wired article

David Merrill from MIT Media Lab shows off his latest creation at TED 2009, programmable computerized play blocks called Siftables. 

The blocks are packed with OLED touch screens, accelerometers, and wireless communications. Merrill and his team have programmed the blocks to do everything from math to cranking out extremely cool chiptunes with a live synthesizer. They’ve also made a killer word game that’s a cross between Boggle and Scrabble, complete with Speak-N-Spell sound effects.

Merrill says the blocks represent a new way to teach and learn, and it’s not hard to imagine them on a glowing plexiglass desk in one of the Enterprise’s classrooms, Data awkwardly arranging them in an attempt to teach a group of wily toddlers. 

I want a set. That word game would be fantastic at parties.

All computer data boils down to ones and zeros. Until now, that is. A team of computer engineers at the University of Pennsylvania have figured out how to throw a “two” into the mix using copper nanowires, adding a third dimension to computing. They call the data triumvirates “trits,” and they could vastly increase the capacity of memory storage devices.

It works like this: Each nanowire is made up of two materials, a central core and a casing. Flashing a current through the wire causes either the core or the casing to phase change from crystalized (neat and orderly) to amorphous (jumbled and messy). The whole wire can either be crystalized or amorphous, representing a one or a zero, a traditional bit. Zapping the core crystalized and the casing amorphous or vice versa, adds the “two,” giving birth to the “trit.” 

Team member Ritesh Agarwal spoke to PhysOrg.com about the discovery:

“The use of nanowires to create electronic memory is advantageous for several reasons, but a non-binary form of nanowire memory like we have created could allow for a huge increase in the memory density of potential future devices.”

That means more memory in smaller packages and, eventually, digital wristwatches that are smarter than I am.

Link to PhysOrg.com article.

Before swarms of nanites can organize to eradicate the human race, they’ll need a leader. Engineers in Japan have made the first steps in creating such a microscopic overlord, building a nanomachine that imitates human brain cells. The tiny machine can receive information from the macro world and transmit it to a small cadre of its companions. Working in concert, teams of the molecular contraptions could do everything from terminate tumors to crunch vast amounts of data in the blink of an eye.

Dr. Anirban Bandyopadhyay of the International Center for Young Scientists, in Tsukuba, Japan, led the team that developed the nanobrain. It’s made from 17 molecules of an compound called duroquinone, 16 arranged in orbit around one. The whole thing is held together by weak hydrogen bonds. Using a scanning electron microscope, Bandyopadhyay was able to send electrical impulses to the central molecule to change its configuration or state. The lead molecule then transfers its state to the other 16, like dominoes falling one after another.

It’s basically parallel processing on a micro scale, the same kind of number crunching that our brains are capable of. In fact, Bandyopadhyay modeled the microbrain on human glial cells, which pass info between neurons in the brain. They call it “one-to-many computation” and it’s key to parallel processing.

So what can it do? Bandyopadhyay estimates that the simple assembly is capable of generating more than 4 billion different outcomes from one input instruction. There’s no comparing true parallel processing to current processors, which crunch computations linearly. Parallel processors can take on millions of lines of instruction at once. That’s the kind of computing power that can keep Moore’s Law of exponential computing growth chugging away into stratospheric heights. 

And it’s not just powerful—the nanocomputer would represent a completely new way of computing. It’s purely visual, using patterns to replace the differential equations that are at the heart of current computing.

There’s also a potential to manufacture billions of molecules of a custom drug with just one instruction. Imagine a single drop of water hitting a placid pool. Waves radiate out from the site of impact, quickly covering the entire surface. A single instruction dropped into a field of similar nanomachines would spread in the same manner.

Bandyopadhyay is currently working to create more complex versions of his nanobrain and hopes to have a functional computer within a few years. The trick is finding something other than a massive tunneling electron microscope to interact with the machines. Bandyopadhyay hopes other control methods will be developed, including optical readers for the nanocomputers, or chemical triggers for the medical nanofactories.

Link to MSNBC article.

Link to BBC article.

Researchers at the Zernike Institute of Advanced Materials at the University of Groningen have developed super-cheap plastic memory that will likely end up in next-gen RFID tags. It works like Flash memory, but it’s easier and cheaper to manufacture. How do they do it? Flash memory is like a club sandwich—layers of semiconductors between ferro-electric toast. The new memory mixes everything up into one blended semiconductor cake. Current can be channeled through the mixture, leaving programming in its wake. The researchers aren’t totally clear on how they’ve managed this feat, but they say it works wonderfully.

Link to ScienceDaily article.

Photo: MRAM never forgets. Ever.

Any technology that employs “spintronics” has got to be aces. MRAM (Magnetoresistive Random Access Memory) exploits the quantum spin states of electrons to store and access data. It’s basically a matter of measuring the resistance caused by different magnetic fields, then making the stuff that generates those fields do what you want it to do. It involves a lot of technical jiggery-pokery, but the end result is RAM that can store data permanently like a hard drive or flash drive, but is as fast as volatile system RAM. MRAM can essentially replace all the memory in a computer—system RAM, video RAM and storage media. Oh, and it uses about 10 percent of the power of DRAM.

The technology has been chugging along on the flat bit of the exponential Kurzweil curve for a few decades now, but it could be ready to shoot into the stratosphere. Several major corporations have decided to dump research money and talent into MRAM and consumers could see MRAM-equipped electronics before 2015.

Freescale Semiconductor, a company that cut its teeth on the venerable Motorola PowerPC chips, announced this week that it will join forces with a few venture capital firms to make MRAM. The new venture, called EverSpin Technologies, will attempt to create MRAM chips that can compete with ever-improving flash memory.

Toshiba and Hitachi are also funding big MRAM projects, as are the U.S. and Korean governments. I can’t wait.

Link to Cnet story.

Link to gizmodo story.

In a move reminiscent of one of the most popular sci-fi plots of the 20th century, the U.S. military is planning to use the world’s fastest super computer to watch over its nukes. The computer, made by IBM, goes by the name “Roadrunner” and uses about 20,000 processors. It runs at ”petaflop speeds,” the equivalent of one thousand trillion calculations per second and is about twice as powerful as the last great super computer, also made by IBM. 

Roadrunner uses the famed “cell” processor, developed by IBM, Sony and Toshiba for the PlayStation 3, to crunch numbers. It employs other standard processors for autonomic functions. Combined, Roadrunner has enough juice to keep an eye on the U.S. nuclear stockpile while simultaneously parsing astronomical, genomic and weather data. The computer will be housed at the Los Alamos National Laboratory in New Mexico.

IBM has several more petaflop computers in the works and plans to make the computers commercially available.

Link to BBC article.

A story’s been bouncing around the blogs about Queen’s University computing professor Roel Vertegaal and the so-called flexible computer prototypes (computers imbedded in a flexible surface like a piece of paper) in his Human Media Laboratory. Here’s the thing: He hasn’t actually made any of the computers. Instead, his team has created a novel projection system that tracks the movement of surfaces and users to create the illusion of a flexible computer. The projectors cast an image on paper (or any surface) as it moves through space. It also identifies the movements and actions of users, allowing them to control the interface through simple movements and gestures.

Vertegaal is testing a user interface based on the assumption that everybody would prefer their computers to be on pieces of paper. Here’s how it works: To scroll trough a document, simply flip the paper compute over vertically. To move from page to page in a browser, flip it over horizontally. To transfer an image or video from one computer (or display) to another, place the computer on top of the other and give it a rub as if you’re making an ink transfer. Voila, the content is transferred. The team demonstrates their interface in a video.

It’s awkward when compared to interfaces being developed by Jeff Han and others. Still, it’s going around, so I’m covering it.

Link to Physorg.com article.