Witt vectors of non-commutative rings and topological cyclic homology (Q1384839): Difference between revisions
From MaRDI portal
Added link to MaRDI item. |
Set profile property. |
||
Property / MaRDI profile type | |||
Property / MaRDI profile type: MaRDI publication profile / rank | |||
Normal rank |
Revision as of 04:09, 5 March 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Witt vectors of non-commutative rings and topological cyclic homology |
scientific article |
Statements
Witt vectors of non-commutative rings and topological cyclic homology (English)
0 references
4 June 1998
0 references
One of the fundamental facts about algebraic K-theory is that the K-groups vanish on a free associative \(k\)-algebra without unit \(F\) over a right noetherian and regular commutative ring \(k\) [\textit{S. M. Gersten}, Commun. Algebra 1, 39-64 (1974; Zbl 0299.18006)]. Hence the K-groups can be viewed as a sort of derived functor. An obvious question is whether the same is true for a given theory purporting to approximate K-theory. When \(k\) is the finite field \(\mathbb{F}_p\), the author proves this for topological cyclic homology \(TC(-;p)\), in the sense that the homotopy groups \(TC_*(F;p)\) vanish in nonnegative degrees. According to \textit{L. Hesselholt} and \textit{I. Madsen} [Topology 36, No. 1, 29-101 (1997; Zbl 0866.55002)] \(TC(A;p)\) is an Eilenberg-MacLane spectrum when \(A\) is an \(\mathbb{F}_p\)-algebra, and since \(TC(A;p)\) is always \(-2\)-connected, the next question must be: how can one best describe \(TC_{-1}(A;p)\)? The author relates \(TC_{-1}(A,p)\) to a group he calls \(W(A)\), which is a generalization to the noncommutative setting of the \(p\)-typical Witt vectors. The group \(W(A)\) comes equipped with additive Frobenius and Verschiebung operators, which reduce to the well known situation when \(A\) is commutative. It turns out that, for any associative ring \(A\), \(TC_{-1}(A,p)\) is isomorphic to the coinvariants \(W(A)_F\), extending the result in [Hesselholt and Madsen, loc. cit.]. The outcome is that \(TC\) of an \(\mathbb{F}_p\)-algebra \(A\) (not necessarily with unit) can be expressed as the left derived functor: \[ TC_*(A;p)\cong L_{*+1}W(A)_F \] Given the close connection between K-theory and topological cyclic homology, the same result follows for the K-theory of finite algebras. The first two sections of the paper are devoted to the definition of \(W(A)\) and its relation to \(TC\). In the last section the author calculates \(TC\) of a free algebra, using calculations and techniques from [Hesselholt and Madsen, ibid]. The paper ends with the interesting fact that whereas the K-groups also vanish on free commutative \(k\)-algebras without unit, the \(TC\) groups do not, even if \(k=\mathbb{F}_p\). One should also note that some restriction on the ground ring \(k\) is necessary for the main result of the paper. Even for very nice \(k\), the homotopy groups \(TC_*(F;p)\) may be nonzero in nonnegative degrees for \(F\) a nonzero free associative \(k\)-algebra without unit.
0 references
topological cyclic homology
0 references
derived functors
0 references
Witt vectors
0 references
Eisenberg-MacLane spectrum
0 references
Frobenius
0 references
Verschiebung
0 references
coinvariants
0 references