Hewlett-Packard claimed this week that technology its researchers have developed may replace the fundamental building block of computers and electronics for the past half-century — the transistor.
Researchers with HP’s Quantum Science Research (QSR) group — who published a paperon the new technology in the most recent Journal of Applied Physics — built a transistor alternativeknown as a “crossbar latch.” The molecular-sized device is capable of providing thesignal restoration and inversion processes for general computing, without the need for transistors, HP said.
The company also claimed the technology could result incomputers that are thousands of times more powerfulthan those in use today.
“We are reinventing the computer at the molecularscale,” said a statement from HP senior fellow and QSRdirector Stan Williams, who is also one of the authorsof the paper. “The crossbar latch provides a keyelement needed for building a computer usingnanometer-sized devices that are relativelyinexpensive and easy to build.”
Tinier Than Transistor
The research is an effort to deal with thelimitations of today’s silicon transistor technology,which is expected to reach its physical limits inabout a decade. In addition to the crossbar latch, theHP researchers are also examiningarchitectural issues and how such tiny devices –thousands of which could fit across the diameter of ahuman hair — can be built in mass quantitieseconomically.
The experimental crossbar latchconsists of a single wire that acts as a signal line, HP reported.It is crossed by two control lines with an electricalswitch junction, allowing control of the device. Whiletransistors can currently handle the same computingtasks, it is generally accepted that they will not beable to shrink to the size of a few nanometers whileremaining operable.
“Transistors will continue to be used for years tocome with conventional silicon circuits,” said astatement from Phil Kuekes, a senior computerarchitect at QSR and coauthor of the paper. “But thiscould someday replace transistors in computers, justas transistors replaced vacuum tubes and vacuum tubesreplaced electromagnetic relays before them.”
Success Gets Smaller
Martin Reynolds, research vice president at Gartner, toldTechNewsWorld that while silicon transistors may have useful lives even beyond 10 years from now, HP’s research was required to head off the limitations of existing transistors.
“At some point, we do reach a point where we can’tmake transistors smaller,” Reynolds said.He added, however, that only time will tell whether the crossbar latch cantruly replace transistors. “At the moment, it’s just a demonstration.”
Reynolds said the trend toward smaller componentsand devices is likely to continue, driven by consumer andcorporate demand.
“For the industry to keep moving, things have tokeep getting smaller, otherwise no one will buyanything,” he said.
Memory Matters
QSR’s Kuekes said the research, partially supported bythe Defense Advanced Research Projects Agency (DARPA),followed demonstrations of working memory andmolecular-scale junctions and represented “the finalcomponent theoretically needed for performing multipleprocessing steps required for useful computing at thenanoscale.”
Reynolds, however, indicated the molecularcomponents used for the technology have slow memoryand can be less stable than traditional transistors.
Nevertheless, the analyst indicated that thecrossbar latch approach would greatly reduce the issueof “noise” or interference that is associated withtransistors.
“It’s neat stuff,” Reynolds said. “It certainlycould become something big.”
Real-World Requirements
Dean McCarron, president of Mercury Research, told TechNewsWorld that although HP’s research may lead tocomputing changes, computers are nonetheless part of much larger systems.
Regardless of the technology they rely on, computers will have to interact with existing electrical infrastructure, architecture and electronics, so transistors will remain crucial.
“It’s one thing to have a computing core based on newer technology, but to get information in and out,you’re going to be relying on today’s technology,” McCarron said.
Loading ...
This is not about PC, Laptop or big chunk of server. HP always had had an AM aizing researcher team and I kind of feel that Mr. Reynold (at Gartner) minimized in my opinion; this breakthrough.
Yes is under lot of research, but HP probe the concept publicaly!!! before anyone.
To: Gartner
Just because wasn’t invented or developed by IBM doesn’t mean is wrong. (For those who doesn’t know founder of Gartner Group is an ex-IBM employee)