Mechanical Memory Switch Outstrips Chip Technology
Mohanty, an assistant professor in BU’s Department of Physics, has carved tiny switches out of silicon, fabricating mechanical switches that are thousands of times smaller than a human hair.
When put through their paces as data storage tools, these nano-sized devices were capable of functioning at densities that far exceed the physical limitations of electromagnetic systems and could retrieve information at speeds that cruise in the megahertz and gigahertz ranges, millions and billions of cycles per second, respectively.
Mohanty also found that the switches operated on miniscule amounts of power, about a million-fold less than that demanded by current systems.
“This is a new ball game,” say Mohanty. “By taking a new look at old technology, we have produced memory cells that are faster and better than those currently used. This mechanical device is a completely new approach to improving data storage.”
via E. Campbell
Interesting. Although, at first, I was confused by the megahertz and gigahertz claims – until it mentioned they were comparing to hard-drive read/write speeds. With DDR2 running at 533-Mhz, for starters, the gigahertz range has already been reached with ‘conventional” technology.
This stuff is neat though, like TI’s DLP chips – with all the micro-mirrors – used in projectors nowadays, and the IBM millipede memory system: While flash memory is not expected to surpass 1-2 gigabytes of capacity in the near term, Millipede technology could pack 10 – 15 gigabytes of data into the same tiny format, without requiring more power for device operation.
But, the IBM info is dated – since 8GB CompactFlash cards were just announced at Photokina! 🙂
Old news, not new, IBM had a release in 2002 about Millipede – http://www.research.ibm.com/resources/news/20020611_millipede.shtml
I think folks miss the point, critical to many designers of portable devices, about energy consumption.
Ed,
Guilty as charged. 🙂
It was the speed and storage-density claims which got my initial attention. I somehow skipped-over the “million-fold less” power requirements – I think mainly because I was trying to determine if it was static-memory, or volatile.