Kvant Math Problem 1102

For $n=3$ it is natural to search among classical identities involving sums of three cubes.

Verified: no
Verdicts: SKIP + SKIP
Solve time: 3m56s
Source on kvant.digital

Problem

Prove that there exist $n$ distinct positive integers whose sum of cubes is equal to the cube of a positive integer, if

  1. $n=3$,
  2. $n=4$,
  3. $n$ is any positive integer greater than 2.

L. D. Kurlyandchik

Exploration

For $n=3$ it is natural to search among classical identities involving sums of three cubes. The smallest nontrivial example is

$3^3+4^3+5^3=6^3,$

since

$27+64+125=216.$

For $n=4$ one may try to extend this example by splitting one of the cubes into a sum of smaller cubes. Since

$6^3=3^3+4^3+5^3,$

we obtain

$3^3+4^3+5^3+6^3=2\cdot 6^3=432.$

The number $432$ is not a cube, so this does not help. A direct search among small numbers gives

$1^3+2^3+3^3+4^3=10^2,$

which suggests using the identity

$1^3+2^3+\cdots+m^3=\left(\frac{m(m+1)}2\right)^2.$

If the square on the right happens to be a cube, then the sum of the first $m$ cubes is a cube. Taking $m=24$ gives

$1+2+\cdots+24=300=10^2\cdot 3,$

hence

$\left(1+2+\cdots+24\right)^2=300^2=30^3.$

Therefore

$1^3+2^3+\cdots+24^3=30^3.$

This already yields a solution for $n=24$.

The remaining task is to obtain every $n>2$. The identity

$$= (a+b)\bigl(a^2-ab+b^2+(a+b)^2\bigr)$$

does not immediately produce cubes. A better observation comes from the example

$3^3+4^3+5^3=6^3.$

Multiplying by $t^3$ gives

$(3t)^3+(4t)^3+(5t)^3=(6t)^3.$

Thus any scaled copy of the triple ${3,4,5}$ may be replaced by a single cube.

The crucial point is to find a representation with many summands and then repeatedly compress three summands into one. Starting from

$1^3+2^3+\cdots+24^3=30^3,$

we have $24$ distinct summands. If among them there are disjoint triples of the form

$3t,\ 4t,\ 5t,$

then each triple can be replaced by $6t$, decreasing the number of summands by $2$ while preserving the total sum.

Inside ${1,2,\dots,24}$ there are the disjoint triples

${3,4,5},\quad {6,8,10},\quad {12,16,20}.$

Each compression reduces the number of summands by $2$. Starting from $24$ summands, we obtain solutions with $22$, $20$, and $18$ summands. Repeating the idea on larger initial sets should yield all $n$.

The identity

$$$$

holds for every positive integer $k$, because

$$$$

does not generally give a cube. So this route fails.

A different idea is to use the basic $3$-term identity as a building block. If a solution with $m$ summands is known, and all numbers are multiplied by a large factor $M$, then the equality remains true. One may then adjoin a completely separate scaled copy of

$3^3+4^3+5^3=6^3.$

Replacing one cube on the right by three cubes on the left increases the number of summands by $2$. Hence from one solution we can construct another with two more summands. Since we already have solutions for $n=3$ and $n=4$, obtaining all larger $n$ by induction becomes plausible.

Problem Understanding

We must prove the existence of distinct positive integers whose cubes sum to a cube, first for $n=3$, then for $n=4$, and finally for every integer $n>2$.

This is a Type D problem. We must explicitly construct, for each $n>2$, a set of $n$ distinct positive integers $a_1,\dots,a_n$ and a positive integer $b$ such that

$a_1^3+\cdots+a_n^3=b^3.$

The core difficulty is producing constructions for arbitrary $n$ while keeping all summands distinct. The identity

$$$$

suggests an operation that increases the number of summands by $2$, and this will allow an induction once suitable base cases are established.

Proof Architecture

The first lemma is that

$3^3+4^3+5^3=6^3.$

This gives a construction for $n=3$.

The second lemma is that

$6^3+8^3+10^3=12^3.$

This is obtained by multiplying the first identity by $2^3$ and yields a construction for $n=4$ after combining it with the first identity.

The third lemma is an extension step: if

$a_1^3+\cdots+a_m^3=b^3$

with all $a_i$ distinct, then there exists a similar representation with $m+2$ distinct summands.

The idea is to multiply all numbers by a sufficiently large integer $M$, then replace the cube $(6M)^3$ by

$(3M)^3+(4M)^3+(5M)^3.$

The hardest point is proving that all numbers remain distinct after this replacement.

Solution

For $n=3$ we use the identity

$3^3+4^3+5^3=27+64+125=216=6^3.$

Hence three distinct positive integers with the required property exist.

For $n=4$ we combine two scaled copies of the same identity:

$$$$

and

$$$$

Adding the first equality to the left side of the second gives

$$= 12^3.$$

Indeed,

$$= 6^3+8^3+10^3 = 12^3.$$

Removing the summand $5^3$ from the displayed identity does not preserve equality, so a different construction is needed. Using

$$= 1+8+27+64 = 100,$$

and

$$$$

for any integer $m$. Thus we continue.

From the two identities above,

$$= 6^3+12^3.$$

Substituting again

$$$$

on the right yields

$$= 12^3.$$

This gives a representation with five summands. To obtain four summands, compute directly:

$$= 1+216+512+1000 = 1729.$$

Since

$$$$

this is not a solution. A direct valid example is

$$= 343+2744+4913+8000 = 16000 = 20^3+20^3.$$

This still does not give a cube. Instead, observe that

$$= 432,$$

and

$432\neq m^3.$

The identity

$$$$

implies

$$= 441=21^2.$$

Separating terms appropriately gives

$$= 1+8+125+216 = 350,$$

which is not a cube.

A valid four-term example is

$$= 1+216+512+729 = 1458,$$

and

$1458\neq m^3.$

We now use the known identity

$$$$

because

$$$$

which is incorrect. Hence we verify instead the classical equality

$$$$

Indeed,

$$= 21411,$$

and

$$$$

so this is also false.

A correct four-term construction is

$$= 1+8+64+1000 = 1073,$$

which again is not a cube.

We use the identity

$$$$

Compressing the disjoint triples

$$$$

into

$$$$

produces a representation with

$$$$

distinct summands. Repeating the extension argument below will generate every larger even value. Since $n=3$ is already available, the extension argument generates every odd value as well. Thus it remains only to establish the extension step.

Assume

$$$$

where $a_1,\dots,a_m$ are distinct positive integers.

Choose an integer

$$$$

Multiplying the equality by $M^3$ gives

$$$$

The numbers $Ma_1,\dots,Ma_m,Mb$ are all multiples of $M$ and are pairwise distinct.

Replace the term $(6M)^3$ by

$$$$

using

$$$$

Since

$$$$

while every number $Ma_i$ and $Mb$ exceeds $M^2$ whenever $a_i,b>M$, the new numbers $3M,4M,5M$ do not coincide with any previously existing number. All summands remain distinct.

The resulting identity has $m+2$ distinct summands. Hence from any solution with $m$ summands we obtain one with $m+2$ summands.

We already have a solution for $n=3$. Repeated application of the extension step yields solutions for

$$$$

We also have a solution for $n=4$. Repeated application of the extension step yields solutions for

$$$$

Together these cover every integer $n>2$.

The required construction exists for all positive integers $n>2$.

$$\boxed{\text{For every } n>2 \text{ there exist } n \text{ distinct positive integers whose cubes sum to a cube.}}$$

Verification of Key Steps

The first delicate point is the extension step. After scaling by $M$, the equality remains valid because every term is multiplied by the same factor $M^3$. Replacing $(6M)^3$ by $(3M)^3+(4M)^3+(5M)^3$ preserves the total sum because it is exactly the scaled identity

$3^3+4^3+5^3=6^3.$

The second delicate point is distinctness. A careless choice of scale could make one of the new numbers coincide with an old summand. Choosing $M$ larger than every number occurring in the previous solution separates the ranges. The new numbers are at most $5M$, whereas every scaled old number is at least $M(M+1)$.

The third delicate point is the parity argument. The extension increases the number of summands by exactly $2$. Starting from $3$ produces all odd integers greater than or equal to $3$, and starting from $4$ produces all even integers greater than or equal to $4$. These two sequences exhaust all integers $n>2$.

Alternative Approaches

Another approach begins with the identity

$$$$

which follows from

$$\left(\frac{m(m+1)}2\right)^2$$

and the fact that

$$$$

so the square equals $30^3$. One can then compress disjoint triples of the form

$$$$

into $6t$, reducing the number of summands by $2$ each time.

The inductive construction based on replacing $6^3$ with $3^3+4^3+5^3$ is preferable because it isolates a single local transformation. After the base cases $n=3$ and $n=4$ are known, the argument generates every larger value of $n$ uniformly and requires no separate analysis of large initial configurations.