Discover magazine has an excellent article, for which I have included a link, that addresses the issue of energy efficiency in computing. As technology proliferates into developing regions of the world, energy consumption is a major issue. A researcher at Stanford University, Kwabena Boahen, has developed a "silicon wafer that provides the basis for a new neural supercomputer, called Neurogrid".
Below is an excerpt that introduces the fundamental shift in design philosophy.
"Traditional digital computers depend on millions of transistors opening and closing with near perfection, making an error less than once per 1 trillion times. It is impressive that our computers are so accurate—but that accuracy is a house of cards. A single transistor accidentally flipping can crash a computer or shift a decimal point in your bank account. Engineers ensure that the millions of transistors on a chip behave reliably by slamming them with high voltages—essentially, pumping up the difference between a 1 and a 0 so that random variations in voltage are less likely to make one look like the other. That is a big reason why computers are such power hogs."
"Scientists have found that the brain’s 100 billion neurons are surprisingly unreliable. Their synapses fail to fire 30 percent to 90 percent of the time. Yet somehow the brain works. Some scientists even see neural noise as the key to human creativity. Boahen and a small group of scientists around the world hope to copy the brain’s noisy calculations and spawn a new era of energy-efficient, intelligent computing. Neurogrid is the test to see if this approach can succeed."
If Neurogrid is successful, System Integration will be taken in a radically different direction, as new applications, and implementations, are introduced. No longer will energy availability be a limitation, and the integration of technology to new markets will fuel innovation.
http://discovermagazine.com/2009/oct/06-brain-like-chip-may-solve-computers-big-problem-energy
Friday, November 20, 2009
Brain-Like Chip May Solve Computers' Big Problem: Energy
Subscribe to:
Post Comments (Atom)
Recent Comments
Archive
Labels
- comic (3)
- introduction (3)
- presentation (2)
- series (3)
- slides (2)
Posts
5 comments:
The concept of nuerogrid is good.But i think it will be very complicated to develop such a thing similar to human neurons.
I feel system integration itself is such a difficult task to fix, implement, run and maintain. And implementing this kind of nuerogrid for system integration might be very very complex....
I agree with Asha that implmenting something as complex as Neurogrid is not feasible as there are enough challenges in the System Integration front. It would take up a a lot of research and training to actually implement this concept in IT. Here is an addition to this term:
"A critical strategy of the NeuroGrid project is to provide enabling capabilities for the neuroimaging community by using Grid technology to allow current algorithms and existing data management procedures to be more accessible and interoperable, so that there are low barriers to entry and time is not wasted re-engineering well established algorithms or forcing alien practices on established research groups. Although the primary aim is not to produce new research data, NeuroGrid will produce valuable new scientific findings from the application of cutting edge analytic methods to combined existing datasets. " Well, it does sound interesting :)
Hey this is totally wacky , but what if somehow we get to use the brain to do the computation , like suppose , we hardly use less than 10% of the brains full potential, then why not use the rest of the brain for computational purposes. Wont it be fun if i could rent a piece of my brain to say salesforce.com , so that they can use the power of my neurons to carry complex calculations.Crazy.But just a thought
It seems that this may not be that complicated. Human neurons are pretty basic mechanisms, but what makes our brain as a whole so functional is the massive parallelism. With the current trend in increasing CPU cores, hardware designers are heading in the same direction. As soon as software able to take advantage of the hardware goes mainstream, we'll be a bit closer to replicating functionality of the human brain. Fun times :)
In a related note. see the link below. Imagine having a SuperComputer that works like the brain and would need to have its various systems integraded with other parts of its network
http://spectrum.ieee.org/computing/hardware/ibm-unveils-a-new-brain-simulator
Post a Comment