TY - JOUR
T1 - Computing in permutation groups without memory
AU - Cameron, Peter Jephson
AU - Fairbairn, Ben
AU - Gadouleau, Maximilien
N1 - Funding: UK Engineering and Physical Sciences Research Council (EP/K033956/1)
PY - 2014/11/2
Y1 - 2014/11/2
N2 - Memoryless computation is a new technique to compute any function of a set of registers by updating one register at a time while using no memory. Its aim is to emulate how computations are performed in modern cores, since they typically involve updates of single registers. The memoryless computation model can be fully expressed in terms of transformation semigroups, or in the case of bijective functions, permutation groups. In this paper, we consider how efficiently permutations can be computed without memory. We determine the minimum number of basic updates required to compute any permutation, or any even permutation. The small number of required instructions shows that very small instruction sets could be encoded on cores to perform memoryless computation. We then start looking at a possible compromise between the size of the instruction set and the length of the resulting programs. We consider updates only involving a limited number of registers. In particular, we show that binary instructions are not enough to compute all permutations without memory when the alphabet size is even. These results, though expressed as properties of special generating sets of the symmetric or alternating groups, provide guidelines on the implementation of memoryless computation.
AB - Memoryless computation is a new technique to compute any function of a set of registers by updating one register at a time while using no memory. Its aim is to emulate how computations are performed in modern cores, since they typically involve updates of single registers. The memoryless computation model can be fully expressed in terms of transformation semigroups, or in the case of bijective functions, permutation groups. In this paper, we consider how efficiently permutations can be computed without memory. We determine the minimum number of basic updates required to compute any permutation, or any even permutation. The small number of required instructions shows that very small instruction sets could be encoded on cores to perform memoryless computation. We then start looking at a possible compromise between the size of the instruction set and the length of the resulting programs. We consider updates only involving a limited number of registers. In particular, we show that binary instructions are not enough to compute all permutations without memory when the alphabet size is even. These results, though expressed as properties of special generating sets of the symmetric or alternating groups, provide guidelines on the implementation of memoryless computation.
KW - Memoryless computation
KW - Permutation groups
UR - http://cjtcs.cs.uchicago.edu/articles/2014/7/contents.html
U2 - 10.4086/cjtcs.2014.007
DO - 10.4086/cjtcs.2014.007
M3 - Article
SN - 1073-0486
VL - 2014
JO - Chicago Journal of Theoretical Computer science
JF - Chicago Journal of Theoretical Computer science
M1 - 7
ER -