P-hard

Arithmetic Circuit Value Problem (ArithCVP)

Input

An encoding $\overline{\alpha}$ of an arithmetic circuit $\alpha$ with dyadic operations $+$, $-$, and $\ast$, together with inputs $x_1, \ldots, x_n$ from a ring.

Problem

Does $\alpha$ on input $x_1, \ldots, x_n$ output 1?

Definitions

See CVP. An arithmetic circuit is a generalization of a circuit which allows doing more arbitrary operations on its inputs.

Status

$\P$-hard, but not known to be in $\P$ (Greenlaw, Hoover and Ruzzo [4]).

Remarks

The reduction to show hardness is from NANDCVP.

The problem is not necessarily in $\FP$ for infinite rings like $\Z$ or $\Q$, since intermediate values need not be of polynomial length. It will be in $\FP$ for any finite ring, and remains $\P$-hard in any ring. It is also $\P$-hard to decide whether all gates in an arithmetic circuit over $\Z$ have “small” values, say values of a magnitude $2^{n^{O(1)}}$. Cucker and Torrecillas [1] look at related problems. The problem is in $\NC$ for circuits of degree $2^{(\log n)^{O(1)}}$, where the degree of a vertex is one for inputs, and $d_1+d_2$ (respectively, $\max(d_1, d_2)$) when the vertex is a product (respectively, sum) of the values computed by two vertices of degree $d_1$ and $d_2$ (Miller, Ramachandran, and Kaltofen [2], Valiant et al. [3]).

References

  1. F. Cucker and A. Torrecillas (1992). Two P-complete problems in the theory of the reals.
  2. G. L. Miller, V. Ramachandran, and E. Kaltofen (1988). Efficient parallel evaluation of straight-line code and arithmetic circuits.
  3. L. G. Valiant, S. Skyum, S. Berkowitz, and C. W. Rackoff (1983). Fast parallel computation of polynomials using few processors.
  4. Greenlaw, R., Hoover, H. J., and Ruzzo, W. L. (1995). Limits to Parallel Computation: P-Completeness Theory. Oxford University Press. Problem A.1.8.