Finite spectra that remove the first chromatic parameters

← All digests

Family 308 · Topology
Importance 5/5 · Provisional editorial rating
Sources consulted 2026-10-22

1. The question

Algebraic topology assigns algebraic invariants to spaces, but not every plausible algebraic object comes from a space or spectrum. A realization problem asks when a desired algebraic pattern has an actual topological model.

Smith–Toda complexes are a particularly sharp example. Brown–Peterson homology has a polynomial coefficient ring with generators $p,v_1,v_2,\ldots$ encoding successive chromatic phenomena. Algebraically, one can simply quotient by the first several generators. Topologically, realizing that quotient by a finite spectrum requires a sequence of genuine attaching maps.

The family claims such realizations at every fixed stage if the prime is allowed to depend on the stage. A second manuscript gives a specified-prime example: an ordinary finite $V(4)$ at $p=1009$. In both statements, the listed parameters are killed to their first powers, not merely to some sufficiently large powers.

2. Background and essential definitions

From spaces to spectra

Suspending a based space adds a dimension in a controlled way. Repeated suspension often makes homotopy problems more stable and algebraic. A spectrum packages this stabilized setting, allowing suspension to be inverted and generalized homology theories to be represented by objects.

A finite CW spectrum is built from finitely many sphere cells by stable attaching maps. “Finite” refers to this cell construction, not to having finitely many homotopy groups or finitely many elements in every invariant.

At a prime $p$, localization makes every integer not divisible by $p$ invertible. The ring

$$\mathbb Z_{(p)}=\left\{\frac ab\in\mathbb Q:p\nmid b\right\}$$

retains $p$-primary information. In these papers, a finite $p$-local spectrum means a retract of the $p$-localization of a finite CW spectrum. A retract is a direct summand in the homotopy-theoretic sense.

Brown–Peterson homology

Brown–Peterson theory, denoted $BP$, is a generalized homology theory related to complex cobordism. At the prime $p$, its coefficient ring is

$$BP_* = \mathbb Z_{(p)}[v_1,v_2,\ldots], \qquad |v_i|=2(p^i-1).$$

The vertical bars give the degree. The $v_i$ are Hazewinkel generators, a standard choice of coordinates. Chromatic homotopy theory organizes stable phenomena through successive parameters of this kind.

For a spectrum $X$, the graded group $BP_*X$ is a module over $BP_*$. It also has a comodule structure over the algebra of cooperations $BP_*BP$. This additional structure records how the homology interacts with the operations of the theory; specifying only the underlying module would leave part of the realization problem unanswered.

A Smith–Toda complex $V(n)$ is a finite $p$-local spectrum with

$$BP_*V(n)\cong BP_* /(p,v_1,\ldots,v_n),$$

including the canonical quotient comodule structure and a generator in degree zero. Fixing that degree rules out silently shifting the entire spectrum.

Killing a parameter requires a map

The cofiber of a map is the stable analogue of attaching a cone to make that map null. If a self-map of a spectrum induces multiplication by a nonzerodivisor $v_i$ on its homology, the cofiber exact sequence gives the quotient by $v_i$.

The algebraic generators form a regular sequence: each next generator is a nonzerodivisor after quotienting by its predecessors. This computes the answer once the self-map exists. It does not construct the map. That gap between algebraic multiplication and an actual topological map is the central difficulty.

3. Development history and previous partial results

Milnor and Novikov established the polynomial structure of complex cobordism coefficients. Brown and Peterson introduced $BP$, Quillen identified the formal-group structure behind cobordism, and Hazewinkel supplied explicit integral coordinates. These results explain the algebra that the Smith–Toda problem asks topology to realize.

The mod-$p$ Moore spectrum, the cofiber of multiplication by $p$ on the sphere, gives $V(0)$ at every prime. Classical work of Adams, Smith and Toda constructed $V(1)$, $V(2)$ and $V(3)$ for $p>2$, $p>4$ and $p>6$, respectively.

Hopkins–Smith periodicity provides generalized Moore spectra at arbitrary finite stages when suitable higher powers of the initial generators are allowed. Those exponents are chosen as part of the construction. Requiring all of them to be one is a stricter problem; the new family does not claim to be the first arbitrary-stage finite realization with freely chosen powers.

The manuscripts cite the September 2023 version of Culver–Zhang, which recorded existence of $V(4)$ as open at every prime. They also distinguish ordinary finite realization from chromatically localized constructions. Kato, Shimomura and Shimomura obtained a related localized object at 1009 with $v_5$ inverted. Inverting a parameter changes the target and does not supply the unlocalized finite spectrum claimed here.

There are genuine nonexistence results at fixed primes. Nave proved that for $p\geq7$, $V((p+1)/2)$ does not exist, and that $V((p-1)/2)$, if it exists, cannot be a ring spectrum. The first obstruction changes its stage with $p$; the second concerns additional multiplicative structure. Neither contradicts existence at each fixed stage with a suitably chosen prime.

Barthel, Schlank and Stapleton developed asymptotic algebraicity through ultraproducts in chromatically localized categories. M1 uses a related large-prime viewpoint but claims to retain ordinary finite spectra before chromatic localization, which is essential to its conclusion.

4. What the new manuscripts claim

Both manuscripts are dated September 23, 2026.

M1 claims all-stage existence with a varying prime. For every integer $n\geq0$, there is a prime $p$ and a finite $p$-local spectrum $X$ such that

$$BP_*X\cong BP_* /(p,v_1,\ldots,v_n)$$

as graded comodules, with the canonical quotient coaction and bottom generator in degree zero.

The order of quantifiers is

$$\forall n\ \exists p\ \exists X.$$

It is not a statement that one prime works at every stage. The theorem does not provide an explicit bound on the necessary prime or assert a multiplication on the final spectrum.

M2 claims a specified-prime realization of $V(4)$. At $p=1009$, it constructs an ordinary finite $p$-local spectrum with

$$BP_*X\cong BP_* /(1009,v_1,v_2,v_3,v_4),$$

with the same comodule and degree conventions. The number 1009 is chosen to make the inequalities in its obstruction calculation sufficiently generous; the manuscript does not claim it is the smallest possible prime.

The two methods are distinct. M1 uses a varying-prime construction and finite descent; M2 works at the stated prime through an Adams–Novikov obstruction calculation.

5. Why the advance matters

Importance: 5/5, provisional. If correct, the family resolves the first-power realization problem at every fixed stage when the prime may vary, and supplies a concrete prime for the first stage beyond the classical $V(3)$ constructions.

These spectra would give finite topological models with especially simple Brown–Peterson homology. They isolate chromatic layers with no extra powers inserted into the defining quotient, making them sharper objects than generalized Moore spectra with larger exponents.

The claim also advances a broader realization theme: a regular algebraic quotient can be produced by actual maps while retaining the cooperation structure, not merely matched as a graded list of groups.

The varying-prime theorem and the explicit 1009 theorem provide complementary information. One reaches arbitrary fixed stages without a numerical prime bound; the other locates a particular realization at a named prime. Neither supplies a classification of all primes and stages where Smith–Toda complexes exist.

6. The main idea

M1: construct a finite diagram across varying primes

M1 forms an ultraproduct of ordinary $p$-local stable categories over the odd primes. An ultrafilter specifies which collections of primes count as large; the construction is not a numerical average of primes.

In fixed homotopy degrees, the resulting mapping objects become rational. But the degrees $2(p^j-1)$ themselves vary with $p$, and discarding that variation would lose the attaching classes. Formal weights retain it.

The paper builds a finite sequence of cofibers, beginning with multiplication by $p$. To construct each next attaching class, it uses a $p$-fold permutation power with a cyclic group and its affine symmetries acting on the factors.

The comparison between a power of a cofiber and the expected cofiber diagram grows with $p$. A central claim is that one uniformly chosen Euler map kills the whole comparison object. An Euler map is a sphere map associated with a representation; inverting it makes the power construction exact in an auxiliary setting.

The desired attaching class must nevertheless be recovered before inversion. Otherwise the result might exist only after localization. The manuscript normalizes the power on universal coefficient algebras, preserving selected nullhomotopies as well as homotopy classes. Connectivity estimates and a finite Koszul construction control the needed comparisons. A Koszul construction is the algebraic mechanism that successively kills a specified sequence while recording the associated nullhomotopies.

Detect the first powers and descend to one prime

Construction and detection are kept separate. Complex cobordism power operations identify the leading term of the newly constructed class with the next coefficient of the formal-group $p$-series, the series encoding multiplication by $p$ in the formal group law.

Passing through the ordinary multiplicative projection to $BP$ identifies the generated ideals with $(p,v_1,\ldots,v_n)$. The argument does not assume that this projection preserves the power operations used earlier.

For fixed $n$, only finitely many attaching maps, cofiber sequences, unit maps and specified homotopies are needed. They descend together to a common large set of primes. Choosing one prime there gives actual finite cofibers in the ordinary $p$-local category.

Regularity then calculates the whole homology module at once. There is no need to intersect infinitely many sets of primes, one for each homological degree. The bottom sphere map determines the canonical quotient coaction and places its generator in degree zero. Full commutative structure in the ultraproduct is not asserted to descend to the final spectrum.

M2: eliminate the obstructions at 1009

M2 first constructs $Y=V(3)$ with a binary product and a two-sided unit. Associativity of that product is not needed for the final step. A sphere map detecting $v_4$ can then be multiplied by $Y$ to give the required self-map.

The Adams–Novikov spectral sequence starts from algebraic data in $BP$-homology and progressively tests which classes come from actual stable maps. A candidate may be killed by a differential on a later page. Thus being an algebraically allowable coefficient does not by itself produce the sphere map.

The paper lists the possible differential targets of $v_4$. A finite filtered calculation reduces them to four cases. The first is excluded by a relation involving the next coefficient and an injective multiplication. Two others are handled through Frobenius-kernel calculations, using the $p$-th-power structure in characteristic $p$ while tracking both polynomial and exterior generators.

The final target requires a topological relation involving a power of the first beta-family sphere class and certain low-filtration classes. Cyclic extended powers and an averaging retract provide the relation. Crucially, the manuscript bounds the page on which the corresponding products vanish. Eventual vanishing in homotopy alone would not rule out the prospective differential at the required time.

Once $v_4$ survives, the spectral-sequence edge supplies a sphere map with the desired Brown–Peterson image. The binary product extends it to a self-map of $Y$; its finite cofiber has the claimed quotient homology. The unit again fixes the comodule structure.

7. Scope, limitations, and open questions

Ordinary finiteness is part of both theorems. A realization only after inverting a chromatic parameter, or a compact object only in a localized category, is not the same conclusion.

The statements kill each listed generator to its first power. They do not replace the established higher-power periodicity constructions, and those constructions alone do not imply these first-power claims.

Neither paper requires a product on the final spectrum. Intermediate algebra or binary products are tools in the constructions, not a promise of associative or commutative ring structures on every $V(n)$. Nave’s multiplicative nonexistence results therefore remain a separate constraint.

M1 leaves the prime unspecified for a general stage. M2 specifies 1009 for $V(4)$ without optimizing it. Determining the exact existence range, minimal primes, and possible multiplicative structures remains beyond the stated conclusions.

8. Sources and further reading

Hopkins–Smith’s periodicity theorem supplies the higher-power background. Compare Smith’s bordism realization work, Toda’s classical spectrum constructions, and Nave’s nonexistence theorem.

Culver–Zhang’s September 2023 version records the earlier $V(4)$ question, while Kato–Shimomura–Shimomura’s localized realization has the different target discussed above. Ravenel’s Complex Cobordism and Stable Homotopy Groups of Spheres develops the coefficient and spectral-sequence background.

Source manuscripts

• M1. OpenAI, Finite Smith–Toda Complexes at Varying Primes (September 23, 2026).
• M2. OpenAI, A finite Smith–Toda complex V(4) at the prime 1009 (September 23, 2026).

Sources consulted on 22 October 2026 at repository revision 1581d4c. Download the manuscript citations.