On Universality and Efficient Arithmetic Realisation of Ternary Logic Gates

Read full text Summarise with VIRAT
Authors: Rajdeep Das, Jyotiprakash Mishra
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

What if the transistor had been built to count to two, not because it was optimal, but because it was convenient? The field of binary (base-2) logic has overshadowed a plethora of alternative possibilities in computing since the 1950s, and certain types of hardware or logical systems may be better suited for specific tasks, a question that is especially relevant in the development of intelligent systems or simulator frameworks. Ternary (base-3) logic is explored as a relatively overlooked yet promising alternative, with potential advantages in information density, circuit depth, and alignment with emerging multi-valued hardware, and its capacity for universality as well as its performance in arithmetic operations as compared to binary systems is examined. All 19,683 binary-arity ternary gates are systematically evaluated, confirming that exactly 3,774 are functionally complete, consistent with Martin’s 1954 result, and the search is optimised via Clone Theory and Group Theory, classifying these gates into 322 equivalence classes under the S3 × Z2 symmetry group and yielding a 47,105.7× search speedup. A key theoretical contribution establishes that unary completeness of a binaryarity gate implies full functional completeness via Rosenberg’s theorem, reducing the universality test from 19,683 binary functions to just 27 unary functions, a speed-up that is 10, 529× faster than the naive approach. Structural analysis reveals that all universal gates are surjective, none are self-dual or zeropreserving, and only 2.4% are commutative. Depth-optimal and size-optimal synthesis algorithms are applied across all 322 class representatives to construct balanced ternary half-adder and fulladder circuits, assessed in terms of gate count, logical depth, and propagation delay against known optimal binary designs. The best-performing gate achieves a half-adder sum depth of 3 and carry depth of 4, yielding a full-adder carry depth of 10 with 29 gates, compared to depth 8 and 21 gates for the binary NAND full adder. Crucially, the information-density advantage of ternary (log2 3 ≈ 1.585 bits per trit) means that at 32 bits, the ternary ripple-carry adder achieves 18% lower propagation depth and 9.4% fewer gates than its binary counterpart, with advantages increasing at wider data paths. Potential implications for emerging computing paradigms, including AI/ML hardware, neuromorphic systems, and analog in-memory computing, are discussed, and a structured framework for evaluating ternary logic in next-generation computational architectures is provided.

Fields of Study & Subjects

semigroups and automata theory Advanced Algebra and Logic VLSI and FPGA Design Techniques

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
31 March 2026
Language
EN
Document type
Preprint