Trinary data storage?!?! It's true!
Researchers from the University of Pennsylvania have created a type of nanowire-based information storage device that is capable of storing three bit values rather than the usual two—that is, "0," "1," and "2" instead of just "0" and "1." This ability could lead to a new generation of high-capacity information storage for electronic devices.
"01001001 01110100 00100111 01110011 00100000 01100001 01101110 00100000 01101111 01110101 01110100 01110010 01100001 01100111 01100101 00100001"
Say data storage units everywhere.
New trinary systems countered merely with "1100220020201010112202022110202"
http://www.physorg.com/news134214217.html
"01001001 01110100 00100111 01110011 00100000 01100001 01101110 00100000 01101111 01110101 01110100 01110010 01100001 01100111 01100101 00100001"
Say data storage units everywhere.
New trinary systems countered merely with "1100220020201010112202022110202"
http://www.physorg.com/news134214217.html
9 Comments
Thats fucking *quality right there. Now I need to learn a new system. Fuck.
Awarding MarineGunrock with one star point for this contribution to Sift Talk - declared quality by arsenault185.
What implications will this have for Siftbot's plans for world domination/subjugation?
so you get four times the space for the same space?
what happens if you can store 0,1,2,3,4,5,6,7,8 or 9?
i am shit at maths.
what are the implications for software? i guess nothing, right? this is just storage...?
^I don't know dick about computation, but yeah, this is just storage. I imagine that any trinary processing is still many years away.
Would information need to be converted back to binary to be read? Wouldn't that increase time to access data?
If nanowire memory becomes the way to go, and is reliably multi-state, then more states is "better". Anything that deviates from base 2 will be a pain in the ass, unless building high-base processors becomes feasible, and has benefits in itself.
Binary systems are simply an abstraction, all memory and processors are actually analog systems which can reliably replicate binary logic by having a large gap of "undefined" values between these states, as you increase the number of states you reduce the amount of undefined space available, and potentially reduce the reliability of the system.
From the link this is physically a 4 state machine, but the difference between states 1 and 2 are undetectable at the moment. If we don't use high-state processors, the high-base memory will create a lot of problems in conversion, even if it has sufficiently distinct states.
Also this is all so pre-alpha, the functional data density is probably many orders of magnitude lower then the HD in my latop, since they probably have less then a meg of the stuff working, and the machine in their lab is probably larger than my fridge.
From a computer tech perspective, it's good to be looking down these roads, but most of them will be dead ends. Don't expect to see anything at this theoretical level ever hit the market.
Thought you might like to know that the crazy Soviets created the first "modern" Ternary Computer, called Setun, in 1958. Apparently they prefer Ternary to Trinary.
http://en.wikipedia.org/wiki/Setun
http://www.computer-museum.ru/english/setun.htm
S4opsUUJ5Ht9IQSJm++z8g2Jy1PYP4uhneo0UF23nrzg5If1D3H5JFO+dz7PRZNSOZeVZXjuhsfxQMdz7bwcffco6lAVmjQNs3zEKGBUGlpHfo666JIgaCCelRGCatZ7
Just had to leave my comment, BTW thats an AES encryption so good luck.
Discuss...
Enable JavaScript to submit a comment.