IBM’s New Computer Chip Thinks Like a Human
Advances in computer technology have become the norm. And as more and more data is collected from a myriad of sources, companies have come to realize the value of using all that information in a multitude of ways. Facebook collects reams of data, and has mastered the art of using it, to better target its ads. And much of the tech sector is racing to both collect and use data to interconnect all facets of our lives — cars, homes, and the workplace, to name but a few — to implement the Internet of Things.
But all of these „smart“ technologies that industry leaders such as IBM (FRA:IBM) are working to bring to the masses require huge amounts of computing power — power that just a few years ago would have seemed unlikely, if not impossible. Driving much of this technological innovation are the capabilities of cognitive computing. In other words,computers such as IBM’s Watson that are able to learn. But even IBM’s supercomputer may take a back seat to what Big Blue calls the world’s first, „neurosynaptic computer chip.“ If „neurosynaptic“ sounds eerily similar to your own brain’s neurons and synapses, this is intentional.
The future of supercomputing?
You might recall that Watson is the now-infamous computer from IBM that beat a couple of longtime Jeopardy contestants. Watson’s capabilities in the world of big data — the collection, analysis, and use of an unprecedented amount of information — is seemingly limitless. And big data offers tremendous potential for IBM and others, including longtime nemesis Microsoft (FRA:MSF) .
Ironically, IBM’s groundbreaking new human brain-like chip was introduced shortly after word spread that the company was trying desperately to unload its money-losing chip business. Talks to sell the chip unit fell apart, but you can be sure IBM’s new SyNAPSE chip wouldn’t have been part of the deal.
IBM’s „brain inspired“ chip boasts an unheard-of 1 million programmable neurons, 256 million synapses, and 5.4 billion transistors, all in a chip the size of a postage stamp. Nearly as impressive as all that computing power is how little energy SyNAPSE requires: a mere sliver compared to common chips.
Traditional chips with a similar number of transistors as SyNAPSE consume as much as 10,000 times more wattage, yet IBM’s product can function much like Google’s famed „neural network,“ which required 1,000 computers, each with 16 processors.
SyNAPSE isn’t going to win any number-crunching contests compared to traditional processors, but that’s not what it’s intended for. Specifically, SyNAPSE and its human brain-inspired functionality are ideally suited for processing sensory data like sights and sounds.
Alone, for now
SyNAPSE uses an utterly unique development process compared to „normal“ chips, and is able to combine the memory and processing functions using all those neurons and synapses, just as with the human brain. As one supercomputing industry pundit said, SyNAPSE „may be a historic development.“ A computer chip that functions much like the human brain? „Historic“ seems apropos.
SyNAPSE puts IBM in a league of its own in terms of how, and what, the new chip is able to accomplish. However, the more broad-based cognitive computing — something SyNAPSE could take to a whole new level in the not-too-distant future — is being employed by Microsoft and other big data specialists.
Microsoft recently upgraded its cloud platform, Azure, to include cognitive computing functionality, directly competing with IBM. Just as with IBM, Microsoft envisions combining the possibilities of big data with the cloud: both have the potential of becoming huge markets. Azure, like Watson and other cognitive computing systems, uses algorithms to improve its results with each iteration, actually „learning“ with each step in the data-crunching process.
Final Foolish thoughts
A computer that learns, such as Microsoft’s Azure and IBM’s Watson, may seem like something from the distant future. But add SyNAPSE into the equation and suddenly a world like The Matrix doesn’t seem quite so far-fetched. How and when IBM can use its new human brain-like chip remains to be seen, but there’s no denying it’s a significant step forward in the world of computing. Who knows: Ten or 15 years from now, when your computer is both answering and asking you increasingly intelligent questions, you can look back to IBM’s introduction of SyNAPSE as the catalyst for a whole new computing world.
Der Bärenmarkt-Überlebensguide: Wie du mit einer Marktkorrektur umgehst!
Ein erneutes Aufflammen von Corona in China, Krieg innerhalb Europas und eine schwächelnde Industrie in Deutschland in Zeiten hoher Inflation und steigender Zinsen. Das sind ziemlich viele Risiken, die deinem Depot nicht guttun.
Hier sind vier Schritte, die man unserer Meinung nach immer vor Augen haben sollte, wenn der Aktienmarkt einen Rücksetzer erlebt.
The Motley Fool recommends Apple, Facebook, Google (A shares), and Google (C shares). The Motley Fool owns shares of Apple, Facebook, Google (A shares), Google (C shares), International Business Machines, and Microsoft.
This article was written by Tim Brugger and originally appeared on Fool.com on 13.8.2014.