Is there chance to make computer with base-3 number system ?

  • Thread starter Thread starter Chitose
  • Start date Start date
  • Tags Tags
    Computer System
Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
6 replies · 4K views
Chitose
Messages
73
Reaction score
0
Is there chance to make computer with "base-3 number system"?

Hello, Chitose wonder chick here again.


No matter how advance cpu these day, It's still base on binary numeral system

0 and 1

So...

I wonder why we can't make computer with base 3 number system?
0 1 and 2 (or what so ever)

I know it's sci-fi but, I've heard something call 'Three diagram',
It's supercomputer that if you give them enough information, It can predict future with near 100% accurate.

..........

English is not my native language, forgive me if I'm wrong in spelling of gamma.
 
Physics news on Phys.org


You could make a computer that was base 3... but why? There is no advantage.
 


At the bottom of it is that 1 and 0 give two states on and off to make computer based on base 3 you would have to determine an intermediate state.
 


For base 3, you could have the state of positive voltage, negative voltage, and 0 voltage as the intermediate step. My gut instinct is this would be less efficient, though it would take an EE person to explain why.
 


I don't think it's simplicity, as much as it is complexity. You have to think more on, how do you store that value physically? I can't elaborate more than that. Someone with more knowledge on semiconductors needs to step in here.
 


There was at least one - SETUN (Russian)

http://en.wikipedia.org/wiki/Setun"
 
Last edited by a moderator:


As TheoMcCloskey points out, it HAS been done, but the other posts that suggest it's likely not advantageous are correct, because of implementation.

I saw a proof once (at least I THINK I remember it being a proof) that the OPTIMUM theoretical base for a compute would be e, but of course no one is EVERY going to build one of those for the same reason ... implementation details make it a bad idea.