Ask HN: Are there any advantages to ternary computing?

12 points by Phithagoras ↗ HN
Reading on Wikipedia the other day, I came across some pages about three-value logic and ternary computing. One of the articles claimed that ternary computers "had notable advantages over the binary computers which eventually replaced it". What could these advantages have been? Also, since there are no ternary computers nowadays, I am curious as to what engineering challenges might have tipped the scales in favour of the binary computer.

5 comments

[ 3.0 ms ] story [ 16.9 ms ] thread
Memristors (if HP and others can bring to production) can support ternary logic.
Can you provide a source on this? I'm not doubting you just haven't heard it yet myself (and was following HP Memristor tech closely).
This is the best Wikipedia page on the topic: http://en.wikipedia.org/wiki/Ternary_computer

Basically, a balanced-ternary system would be smaller and more efficient than a binary system that can match it. I'd like to see an open architecture chip designed and built with balanced ternary logic. It would probably need to be made from memristors once that is feasible. We will need a computing reboot at that time anyway in order to take full advantage of memristors.

Somewhat ironic (and self-answering) asking that question as a boolean..
The advantages of using ternary logic vs binary logic to do our computing would be that we can hold the same amount of information in less memory, and we can process more information at once.

To implement the hardware based on ternary logic is much more complex and that is why binary won. It is much easier to process and store 2 representational states in each bit then 3. It is easier to tell if there is a charge or not vs determining the difference between a high charge, a medium charge, or no charge. It is easier to store and read information with 2 polarities vs 3 polarities.