Multiplicities and Betti numbers in local algebra via lim Ulrich points (Q2166206)

The present paper concerns multiplicities of modules, and finite free complexes with finite-length homology over a commutative Noetherian local ring \(R\). One of the questions they investigate is that: For a given local ring \(R\), what is the smallest possible length of a nonzero \(R\)-module having finite projective dimension? The authors conjecture that (Conjecture 1) \(\ell_R( M) \geq e(R)\) for every nonzero \(R\)-module \(M\) of finite projective dimension, where \(\ell_R( M)\) (resp. \(e(R)\)) denotes the length of \(M\) (resp. the Hilbert-Samuel multiplicity of \(R\)). Conjecture 1 implies the still-open Lech's conjecture, at least for Cohen-Macaulay rings. This measures the difficulty of this conjecture. Under certain assumptions on \(R\), they obtain a lower bound on \(\ell_R (M)\) in terms of \(e(R)\) for every \(R\)-module \(M\) of finite length and finite projective dimension. As a consequence, they prove Conjecture 1 for strict complete intersection rings. Next, they obtain a lower bound on the Dutta multiplicity of \(M\), denoted by \(\chi_\infty(M)\), when \(R\) is the localization at its maximal ideal of a standard graded algebra over a field of positive prime characteristic and \(M\) is a finitely generated \(R\)-module \(M\) of finite length and finite projective dimension. This result connects to earlier result in the case \(\chi_\infty(M)=\ell_R (M).\) This holds, for instance, when \(R\) is a complete intersection ring. More generally, they obtain this lower bound for non-Cohen Macaulay rings. In this case, every nonzero module of finite projective dimension has infinite length. So, they focus on ``short complexes supported on the maximal ideal'', by which they mean that a nonexact complex of finite free \(R\)-modules of length dim \(R\) such that all homologies of this complex have finite length. Inspired by this result, they conjecture (Conjecture 2) that \(\chi_\infty(F) \geq e(R)\) for every short complex \(F\) supported on the maximal ideal. They prove that Conjecture 2 implies Lech's conjecture. The key idea in the proof of these results is the construction of a lim Ulrich sequence of \(R\)-modules. Ma has shown that every standard graded algebra over a perfect field of positive characteristic admits a lim Ulrich sequence of modules. In this paper they revisit the construction of such sequences from a geometric point of view. In particular, they introduce the notion of a ``lim Ulrich sequence of sheaves'' on a projective scheme over a field. They establish that every projective scheme over an infinite, perfect field of positive characteristic \(p\) admits a lim Ulrich sequence of sheaves. They recover Ma's construction of lim Ulrich sequence of modules over such rings.