Digital Quantum Computing with algorithms

By Ed Gerck © 6/26/2024.
egerck@gmail.com
PlanaltoResearch
781 Washington St. #3423
Sonora, CA 95370
ORCID: 0000-0002-0128-5875
SCOPUS ID: 6603293641

MSC: 68Q12. Keywords: Quantum and Classical Artificial Intelligence;

Schrödinger equation; large prime numbers; factorization; quantum
computing; high temperature; cell phones; Linux; eigenvalue;
diophantine algorithms; congruence.

ABSTRACT
Digital quantum computing (QC) uses algebra and quantum mechanics.
After 42 years of frustrating analogue efforts, we dismiss the influence of
time, materials, temperature, and delocalization. The Schrödinger Equation
(Equation) in all cases presents a well-known eigenvalue algorithm Hψ=Eψ.
An AI model, particularly suited to machines, reveals trillions of new logic
equations. This discovers only integer eigenvalues forming diophantine
algorithms, solved by congruence relations; reaching digital QC algorithms.
This makes the sets N and Q rigorous classifications in AI; while all
observations in nature are harmonized with zero model error. Eigenvalues
are physically observable as generalized resonances including quantum
superposition and entanglement, in an algebraic-numeric computation using
values only in the set B={0,1}. Digital QC solves how to do quantum
computation. This is confirmed by the Equation's potential V(M, x) (i.e.,
Physics), providing all the Equation's eigenvalues. The observable values
using V(M,x) are given by the first M prime numbers, and connects with the
AI model in a finite field in mathematics. Digital hardware is poised to use
switches in commercial tri-state chips, and FPGAs. We hope cellphones
become ubiquitous high-performance supercomputers. We discover that RSA
is mathematically broken. Non-Target Cryptography is recommended.

1. Introduction

1
Theoretical and experimental integer factorization is well-known to
have been for a long time the exclusive domain of Number Theory in
pure Mathematics. It is also well-known that searching for a formula
or pattern that can predict prime values has not been successful so far
in Mathematics. This unpredictability adds an element of mystery to
prime numbers, making them a tantalizing subject of study.

Prime numbers are also well-known to play a pivotal role in the field
of Cryptography, which follows Mathematics. The factorization of
any integer values (large or small) is expected algebraically from the
Prime Factor Theorem (PFT)1 since 1873. Algebra does not depend
on value length.

In other words, a prime number, with 1 excluded, is divisible only by

itself; every odd value > 2 is a prime value, or a composite number
with a prime value; and there is no false positive in the algebraic
view, nor integer length limitations.

This breaks the well-known method of encryption described by RSA,

for any particular RSA modulus (i.e., a product of two prime
numbers), or "strength" as integer length. Factoring the product of
two large prime numbers is a direct consequence of the PFT, a
mathematical fact at any integer length, but that has been ignored
blithely in Cryptography. This directly breaks all existing public-key
algorithms (RSA), creating challenges and is well-known to evoke
chaos. Breaking RSA is expected, though, by the US NIST (using
undisclosed reasons).

This contradiction must create an improvement in cybersecurity

policies by using mathematics already known since 1837. This also
avoids the old and ineffective cybersecurity practice of obfuscation,

1
The PFT, by Gauss in 1873, states that every integer greater than 1 can be uniquely
represented as a product of integer prime numbers, irrespective of integer length and
factor order. This has been well-known for more than 150 years.
2
although well-known and reasonably effective before computers were
machines, introducing AI2.

Why should Physics impact these fields?

Physics discovered that everything we thought was physical and

continuous is not … according to quantum mechanics (QM). This
"came out of the left field", unexpectedly, into mathematics,
contradicting the continuity of values.
In QM, numbers must be isolated. This involves two aspects: isolation
and separation. Any two quantum numbers must be separated from
each other, and not be in contact. In consequence, they must be exact,
as a point of dimension 0. In a mathematical finite field, this maps 1:1
with natural numbers and rational numbers as values.

Matter has been revealed to be mostly empty space and matter can
change into energy —- and vice-versa, according to the well-known
E/c2 = m.

These new and unexpected results in Physics have turned quantum

mechanics (QM) into quantum computing (QC), and revolutionized
expectations with a QM understanding of values -- where they must
all be discrete (i.e., values must include the quantum aspect). QC is
well-known to have emphasized discrete mathematics, with modular
arithmetic.

QC has been reported using mathematical, digital algorithms most

recently in 2023 [1], with the differentiation of discontinuous
functions -- well-known to be prohibited by current Mathematics.
Within the past 42 years, there were several attempts to use analog
techniques in the factoring of values, with Shor's algorithm [2-4],
adiabatic QC [5-10], quantum annealing principles [10-11], and
others [12-14]. The largest number factored so far by these techniques

2
While AI and quantum evidence of the distribution of prime values are presented
here, and used also to foresee large prime values, a further exploration of the topic is
to be presented elsewhere.
3
[2-15] is only 4088459 [15], which is very small for cybersecurity
applications. Earlier attempts [16] using calculations in optical
interferometer cavities, could not present sufficient resolution for
higher modes: using a 100x100 matrix, one could calculate typically
less than 10 dominant modes and about 90 random collections of
points, due to the physical non convergence of the process. The same
seems to affect analogue QC using materials, not optics. Scott
Aaronson summarized the view of analog QC in [17]. The potential
consequences of QC in creating panic has been summarized in [18].

The understanding originated from Physics in QM and QC has not

been shared yet by conventional Mathematics and Cybersecurity, with
a well-known Not Invented Here (NIH) Syndrome that has separated
natural sciences from deductive, logical sciences. This creates an
incentive for this work -- to review and share the significance of QM
and QC to so far missing applications in science3.

As another motivation for this work, and ultimately accretive, one can
refer to the well-known "Holographic Principle" in revealing
connections between Physics, Mathematics, and Computer Science.
In conventional (e.g., classical) computers, as well-known, a bit can
only be in one state at a time, either a zero or a one. If we want to do
a (classical) calculation, we could only do it for one state at a time
only. Not so, QC promises.

QC uses internal variables called qubits [2-15]. Qubits also consist of

two states, zero or one, but unlike a classical bit, a qubit does not have
to be in just one state at a time. It can be in an arbitrary combination
of those states, of quantum superposition and entanglement, if one
could so describe it, of zero and one.

If one has two qubits, they can exist simultaneously in a superposition

of four states. When one repeats a calculation done classically, such
as multiplication by 7, one can actually perform the calculation for all
of those states at the same time, and one is left with a superposition of
3
The feeling of missing out (FoMO) has been real.
4
the different answers: 1, 7, 49 and 343. By adding another qubit, one
doubles the number of possible states. Increase that number to just 20
qubits, and one can already represent over a million different states,
meaning one can simultaneously compute over a million different
answers. With 300 qubits, one can represent more logical states than
there are particles in the observable universe4.

This sounds incredibly powerful, but it is not magic. There is one very
big catch: all of the answers to the computation are embedded in a
quantum superposition of states. One is not expected to simply read
out this superposition as observable quantities.

When one makes such a physical measurement, the only observable

value seems single -- i.e, one value from the superposition, basically
at random, and all the other information is lost.

Thus, in order to harness the power of a quantum computer, one needs

to first find a smart way to convert a quantum superposition of states
to the one which contains only the information one wants.

This is considered an incredibly difficult task [1-17] in QC, and is the

“QC question” in [17] and here.

That is why, for most applications, quantum computers have been

useless. So far, [2-15] have only identified a few problems where one
can actually hope to do this significantly, albeit for a small number of
bits.

It is, thus, well-known that QC, in any form, cannot produce qubits,
which are not physically observable. Even with additional 'work'
qubits (i.e., it is well-known that at least 3 logical states are needed),
the ensemble of qubits collapses to a ... digital value -- "poof" -- when
physically measured. The well-known Schrödinger cat is eventually ...
dead.

4
Maximum is estimated to be about 1080 .
5
Only a digital value (exactly 0 or 1) seems physically observable -- as
unique values, albeit with arbitrary names. No one has been able to
measure an observable mechanism using the final configuration of
qubits, nor that Schrödinger's cat would be (somehow) magically
alive while dead.

The √−1 (symbol "i") is also not physically observable, though used
in analog QC. No one ever saw an imaginary value in nature -- phase
can be accounted for by rotation, e.g. , in well-known holographic
calculations and in electric reactive current.

QC cannot directly produce, thus, complex and imaginary values. Nor

produce irrational numbers and mathematical real-numbers, both with
infinite information value as mentioned by the author and Gisin [1,
19] and not used by the author in [20] -- they are not physically
measurable, nor ever seen.

The algorithms in a mathematical, digital form of QC can, however,

use well-known arbitrary-length (AL) integers to calculate without
any programming limitations. AL is called BigInt in most
programming languages, and is available in the well-known program
bc in Linux.

The “QC question” is solved in this work in Section 2.2. This solves
the fundamental problem of QC -- the measurement of qubit patterns
in a smart, unique, physically observable way, which can be used as
objective values.

The solution (see Section 2.2) is that qubits are physically observable
as eigenvalues in the set N, and present their patterns as generalized
resonances including superposition and entanglement -- which are
exact and measurable in their final state, in an algebraic-numeric
manipulation using only what computers can use, the set B={0,1}.

This work describes this stunning solution, working today, reaching

10991 states [21] and more, showing numerically the calculation of AL

6
prime numbers, non sequentially, with no physical limitations and
with no time, material, delocalization, or temperature dependance.

The solution has ten steps: The Schrödinger Equation as described in

Sec.(2.1); the resulting eigenvalue algorithm at any temperature (e.g.,
pocket temperatures) as described in Sec.(2.2); a new Quantum and
Classical Artificial Intelligence is described in Sec.(2.3) forming the
AI model; discovering new isomorphisms as described in Sec.(2.4);
with only integer eigenvalues as described in Sec.(2.5).

This allows us to formulate all problems with diophantine algorithms

as described in Sec.(2.5); find new congruence relations as described
in Sec.(2.6); and discover QC as described in Sec.(2.7).

As a result, one can use the sets N and Q in the AI model in rigorous
classifications as described in Sec.(2.8); which allows all observations
in nature to be harmonized with zero error in the model, as described
in Sec.(2.9). Any variance does not propagate and change, as every
process is exact, albeit stimulation out of range is included.

Digital QC solves in the final step, as described in Sec.(2.10), the

“QC question”, where we show how to do quantum computation with
qubits in patterns (values) that are observable by anyone equally
(objective).

Sections 2.1-2.10 are confirmed by the Equation's (prime number)

potential V(M, x) (i.e., Physics) providing exactly the Equation's
eigenvalues (i.e., Mathematics), which are the observable solutions as
prime numbers -- given by the first M prime numbers, and connect
with the AI model in a mathematical finite field.

We further show in the Discussion, that these steps can be a new

foundation for Cryptography as a science, away from the public key
cryptography that relies on obfuscation, as used today. We are
proposing No-Target Cryptography (NTC).

2. Materials and Methods

7
As computing means we used a commercial Linux distribution, in a
commercial hardware, with the software bc providing AL integer
operations, and capable of running C and BEND code, with C version
12. We also used a commercial cellphone when possible5.

As to the method, we describe it in Sections 2.1 through 2.10, next.

2.1. Schrödinger Equation

This topic is well-known, as in [22]. In summary: each QM

observable is determined by and solved by an eigenvalue problem
Hψ=Eψ with different operators H for different observables; if a
specific function is an eigenfunction of a specific operation, it can be
used to extract the corresponding observable, which is the eigenvalue
E; the Schrödinger Equation (Equation) is well-known to be an
eigenvalue problem with the eigenvalue ascribed to total energy E.

This solves in a simple way the QC question [17] of reading

information patterns encoded in qubits: extract the qubits using the
eigenvalues, and this is exemplified numerically in [20-21, 25].

Both time-dependent and time-independent Equations are the best

known instances of eigenvalue equations in QM, with their
eigenvalues corresponding to the allowed energy levels of the
quantum system [19]:

-ℏ2/2m[∇2+V(r⃗ )]ψ(r⃗ ) = Eψ(r⃗ ) (1)

The object on the left that acts on ψ(x)is an example of an operator

[19]:

−ℏ2/2m[∇2+V(r⃗ )] (2)

5
The US Moon program in the 80’s used 3 IBM 360 J, and a team of technicians in a
refrigerated room, with far less capacity in memory and operations per second than a
commercial cellphone is capable of, in one’s pocket today.
8
All physical models using the Equation can be cast as an eigenset
algorithm:

Hψ=Eψ, (3)

where H is an operator.

2.2. Eigenvalue algorithm

Solving eigenset algorithms such as Eq.(3) is discussed in linear

algebra. Particularly, we will cast the eigenvalues (which are
calculated as AL integers in Section 2.4) as a series E= a + bn + c/n +
d/n2 + … as done in [20], or as a continued fraction as done in [25].

We see from [20, Eq.(3)] that only rational eigenvalues appear when
the Equation is piecewise discretized in exactly three points, solving
all given potentials V(r⃗) in the Equation to any finite degree of
expansion, e.g. using matrix representations of AL order [25]. In
the matrix model [25], each eigenfunction is a piecewise linear
combination in the set {e-x, xe-x, x2e-x}, x in the set Q, and the
eigenvalues are in the main diagonal (irrespective of the AL
discretization used).

Considering any AL discretization, a finite set of rational eigenvalues

results. As well-known, all such members of a finite set of rationals
can be written as integers. Let d be the largest such integer, where d is
the Greatest Common Divisor (GCD) of the set, which leads to a
finite set of integer eigenvalues.

One can use the same GCD procedure in a finite series or in a finite
continued fraction expansion -- calling both expressions “open-ended
integers” (OEI) to represent a finite set of integer eigenvalues in each
format. Other formats can be considered for the eigenvalues, albeit
always as OEI in a finite set, for any potential in the Equation.

Insight comes before application, as a well-known quote by Max

Planck. The insight of ‘qubits → eigenvalues’ harnesses the unique

9
power of a quantum computer, where one has found through physics a
proven and smart way to mathematically convert a quantum
superposition of states to the one which contains only the information
one wants -- the eigenvalues seen as generalized resonances of qubits,
expressed exactly as OEI, in a finite set albeit with AL. This also
replaces the informal and unquantifiable notion of “infinite”.

Each eigenvalue depends on all the others, as well-known in linear

algebra. When one uses the Equation's potential V(M, x) (i.e.,
Physics) providing the Equation's eigenvalues (i.e, Mathematics),
which are given by the first M prime numbers, and connect with the
AI model (in Section 2.3) in a mathematical finite field; this is
evidenced in the tri-diagonal form of the operator H as the main
diagonal of the discretized operator H [19] in any AL, showing as
eigenvalues all prime numbers up to M, and no other value.

One can also use the format of the eigenvalues to easily define
patterns; for example, if the eigenvalues for V(M,x) obey E = an + b
(like the harmonic oscillator eigenvalues (e.g., as bosons are modeled)
[20], any two eigenvalues will be separated linearly. Compared with
the same expression in [16], the quantum property in Eq.(4) below is
also evidenced by “eigenvalue translation”, keeping the eigenvectors
constant in the solution:

E = 2n + C, (4)

where C is a prime number, and n is in N. Because a prime number

larger than 2 is always an odd value, while an odd value plus an even
value is always an odd value, this also results in Eq.(4), confirms it,
and hints at the quantum properties we find in integers (see Section
2.3).

With C = 3 and n > 5, one has to numerically eliminate powers of the

previous prime numbers, which is done by expressing them as
exceptions as: \x. This is, unfortunately, a sequential and slow method
to eliminate false positives that appear in the numerical view, albeit

10
the quantum equation Eq.(4) never produces a false positive in the
algebraic view (the value is either in a composite number of prime
numbers or a prime number by itself). This leads us to consider a
short-cut to break RSA.

According to Gauss, because each RSA modulus is a product of two

prime numbers (albeit unknown), looking for the two prime numbers
in an RSA modulus is exact in algebra, not a brute search, not
random, not incalculable. Let's accept that as the proven truth, in spite
of RSA blithely ignoring it.

We can now use a non sequential, fast, quantum method at inception

using Eq.(4). The expression used is reserved for publication
elsewhere. Readers can informally already forecast a large prime
value C, without needing a sequential search -- which saves some
trillions of years over classically computing prime numbers.

2.3. AI model

Using the quantum patterns, and others, defined in Section 2.1, we

form a new Quantum and Classical Artificial Intelligence model
(called our AI model).

To create a non-human, larger, space in logic, our AI finds statements

that machines can explore methodically, tirelessly, and expand further,
while confounding humans.

Our AI should use this effectively while humans struggle to use it

correctly, creating a machine space mostly complementary and useful
to humans.

Our AI uses billions of new logic laws! Such as governing

non-boolean logic. Our AI can also use AL integers, easily.

This expands to literally trillions of new, already valid statements.

11
The logical states are also in different dimensions, and as well-known,
a continuous path in a higher dimension must necessarily map into a
discontinuous path in a lower dimension -- so disconnected points
(the quantum effect) are seen -- not continuity.

Any possible errors or faults (e.g., alpha particles) are accountable for
by enough redundancy. In general, as well-known, a triple system
solves it by internal coherence.

This was used in India in the Mogul period, to effectively access fair
taxes.

One example in our AI is contrapositive logic. This is based on the

fact that a solution ALWAYS exists! Particularly when unknown.
When solving for "x" in algebra, the first hypothesis is that "x" exists!

Most comments on this work do not see that "contrapositive logic" is

correct and easy to apply, and even go to lack of logic in order to,
somehow, find a fault.

"Contrapositive logic" is seldom used by humans. We notice, further,

an avoidance in using it.

Humans that still believe in "disasters" as both existing and

inevitable, need to learn it. In quantum terms, everything we thought
was continuous and material, are not. We no longer have to fear the
unknown, or gaps, as this work shows logically.

Contrapositive logic means a proposition or theorem formed by

contradicting both the subject and predicate or both the hypothesis
and conclusion of a given proposition or theorem and interchanging
them.

E.g., if "P, then Q" is equivalent to "if not Q, then not P".

Or, "If it is raining, then I wear my coat" — "If I don't wear my coat,
then it isn't raining."

12
The law of contraposition says that a conditional statement is true iff
(if and only if) its contrapositive is true.

That all known cybersecurity systems today are not 100% secure,
becomes actually positive.

It means that at least an unknown cybersecurity system must be 100%

secure.

We will publish a candidate elsewhere. "Impossible" results will

follow, as if “coming from a left field" -- unexpectedly.

Claude Shannon did the same with his 10th theorem -- showing that
"zero error" actually can be reached, not just "approaching zero."
This enabled today's error-free solutions in storage, and RAID
systems.

That all known cybersecurity systems today are not 100% secure,
means that at least an unknown cybersecurity system must be 100%
secure.

Failure all-around does not mean that one is doomed to more failure --
but must "seek something else". A solution must exist.

If a target exists, it will certainly cause a failure mode, which means

that to "not have any failure modes" is indicated, exemplified by what
we are proposing in Sect.(4), as No-Target Cryptography (NTC).
NTC includes an internal measure of success, using coherence, and is
fault-healing by using enough redundancy of independent systems, as
well-known.

When making binary decisions, one seems helped by the LEM: Law
of the Excluded Middle. This is fast. However, this is recognized as
an internal measure of success, but artificial, much like a Procrustean
Bed. In the routine evaluation of going or not to the beach, a Boolean
decision that can benefit from the LEM does favor a fast decision, but
one has to face an estimate of many non-Boolean variables. Weather,
food availability, schedule, proximity, trip considerations,
13
neighborhood, access, and other factors routinely play an important
role, and may be difficult to estimate, oscillate, and have differing
intersubjectIve interpretations. The solution has been to enlarge the
system under consideration, but no system seems to be ever complete
enough, and this can only lead to a probabilistic model.

Instead, to reduce the well-known resulting irreproducibility, our AI

model uses a non-boolean model, with many finite values, suitable for
use by machines.

A similar situation is observed in mathematics, physics, and any

natural science. In biology, e.g., in investigating melanocytic lesions,
one is well-known to be faced with a binary decision (e.g., as to
whether a lesion is benign or malignant) but one has to consider
non-Boolean variables. This can lead to a 99.99% probability using
simple, well-known Bayes statistics in repeated tests. A very small
(0.01%) variance may not exist to accommodate any error as
well-known, but may express variations in the environment.

These are all examples of a final Boolean decision, with only two
wanted final states: go or not go, benign or malignant. However, in
approaching this problem, one has to apply a number of criteria for
this interpretation, since no single criterion seems to be sufficient at
start.

In particular, we find at least 4 quantum properties that are always

present in the number systems of sets N, Z, and Q. This is explored in
[1], where the first three quantum properties are discussed. The fourth
quantum property of integers can be seen here in prime numbers, as
indistinguishability.

Indistinguishability sets in because one cannot distinguish any digit in

the prime number 163283381 (as verified by a cellphone calculator).
It is a “blob” and cannot be split in any way. The same happens in the
prime number “11” , where the digit “1” cannot be distinguished.
These prime numbers are valid, using the set N, here on Earth and in
the star Betelgeuse. It is a universal quantum property, the fourth, and
14
can be used for communication between entities that have never met
before.

This work deals also with this problem, without taking a probabilistic
approach, or proposing immeasurably small values. This may lead
one to deprecate the well-known Gödel's uncertainties, and solve the
liar's paradox as differing intersubjective interpretations,
corresponding to different classifications in the AI model, which is to
be published elsewhere. Next, using the foregoing, we create the
Quantum and Classical Artificial Intelligence (AI) model.

AI is a bit of a misnomer. AI means processing logically in a device

or software the inputs, and providing outputs that can be external
and/or internal (memory).

The logic in AI provides correspondence to categories using

intersubjective processing provided by humans.

In short, humans control the intersubjectivity "button" providing the

metadata to interpret the data (is it a child or a tree?) and AI logically
(and blindly) executes the data following the metadata.

AI cannot provide yet, reliably, the intersubjectivity decisions. AI

may in the future provide also the metadata in addition to the data, for
example using identification, trust, memory, risk analysis, risk
avoidance, and profiles. (e.,g., mud vs. ice).

The AI model can be set to optimally search numerically for a pair of

prime numbers p and q, to factor any RSA modulus as p.q. The
existence of such a pair is assured mathematically by algebra --
obeyed by any RSA modulus, well-known and discovered by Gauss.

The prime numbers p and q do not have to be found by sequential

search as in a classical computer running the Generic Number Sieve,
starting from 2. This, as well-known, should take trillions of years and
is considered practically impossible.

15
Using the AI model after digital QC, prime numbers can be formed as
generalized resonances using superposition and entanglement, e.g.,
eigenvalues, exactly, and read arbitrarily, without any obligatory
sequencing.

Prime numbers also become more regular with larger length: it is

well-known that the likelihood that a randomly chosen number less
than n is prime is (approximately) inversely proportional to the
number of digits in n.

This makes higher prime numbers more predictable, which also

supports the quantum Eq.(4) numerically.

The AI model uses several well-known algorithms, such as binary

search, the Rabin-Miller test [23], and others, including a three-point
propagator (see Section 2.1) solving all potentials in the Equation.
Eigenvalue translation [16] uses an eigenvalue estimation technique
from linear algebra (including the well-known Prony method). Using
these algorithms and the known modulus K, digital QC can produce a
prime number p’ to 0.2% variance or less. This can be verified
deterministically by using one division in AL integers producing a
candidate q’.

The AI model and digital QC proceed to reach deterministically the

unique p.q that reproduces numerically the desired RSA modulus.
This breaks RSA, and can be used legally by law-enforcement, court
order, to combat well-known cyberpiracy, and by accredited
organizations or persons, as NMA is developing.

2.4. Integer eigenvalues only

As is well-known, every integer value corresponds to one or more

rational values, but rational values do not always correspond to
integer values. However, for finite sets, a well-known isomorphism
exists as a 1:1 map between integer and rational values.

16
Therefore, we discover integer eigenvalues only, as observable
solutions of the Equation's eigenvalues given by the first M prime
numbers, solving the Equation's potential V(M, x) (i.e., Physics). This
connects with the AI model in a mathematical finite field (i.e.,
discrete Mathematics).

2.5. Diophantine algorithms

The previous Section 2.4 leads us directly to diophantine algorithms

-- typically using a polynomial equation with integer coefficients, for
which only integer solutions are of interest.

2.6. New congruence relations

The well-known Euler method for solving a diophantine equation

involves taking congruences, and also changing congruences to
equations. As well-known, a ≡ b mod n, means that n divides (a − b)
or a − b = nt for some t, or a = b + nt. This is useful to be able to
convert a ≡ b mod n to a = b + nt for some t. Conversely, if one has an
equation a = b + nt, one can use mod n, remembering that n ≡ 0 mod
n. So, from the form a = b + nt one gets a ≡ b + 0t mod n, or a ≡ b
mod n.

As a new congruence relation to find a prime number, we use the

quantum Eq.(4). This congruence relation can skip all previous prime
numbers (i.e., as a quantum jump) allowing us to skip trillions of
years of calculations, and landing on the desired prime number,
leading directly to a 2046 bit modulus.

One prime number also does not perfectly mask the other prime
number in multiplication in the RSA modulus, because they cannot
have the same length. The longer one is revealed verbatim in the
leftmost bits (See Section 2.7).

To find the desired prime number deterministically (i.e., as good as

the RSA modulus in question), we use the new (created in 2023)
computer language BEND [24]. BEND is a massively parallel,

17
high-level programming language, based on Rust and C. It runs on
massively parallel hardware like GPUs, with near-linear speedup
based on core count, and zero explicit parallel annotations: no thread
spawning, no locks, or other care.

2.7. Digital QC

The above has described an algorithm using diophantine equations.

The algorithm implements QC digitally, to find prime numbers in
breaking numerically a RSA modulus. With enough resolution to
show cryptographic value, such as in numerically breaking
RSA-2048.

This works with integers only, with no temperature effects. This uses
algebra and analytic methods as well as numeric computation, where
digital QC solves otherwise impossible problems, with generalized
resonances using superposition and entanglement, e.g., eigenvalues, --
exactly, and measurable, as an algebraic-numeric calculation using
only what computers use -- the set B={0,1}.

The unique prime number that breaks the well-known RSA-2048 can
be calculated in an AL integer operation, such as provided without
cost by bc in Linux, or in C code.

This seems impossible because multiplication seems like a trap-door

function. It is easy to calculate p.q=K but seems impossible to
calculate p and q given K.

Not with QC. Not even with algebra -- with the well-known Prime
Function Theorem found by Gauss 150 years ago. There are 2 unique
primes in any case of an RSA modulus.

To break RSA, one just needs algebra, and find one prime number.

QC justifies this. But arithmetic can do it. AI helps. But it also works
without AI.

18
A demonstration:
RSA is broken by algebra (Gauss, cited PFT) and algebra never
depends on the length of numbers.

Length of numbers only adds obfuscation. 7×5 = 35 is enough. How

safe is 7 or 5 in 35?

But we can know those values, and not even with QC!

Given K, in general, a unique p and q do exist, but we can't

numerically find them using a human calculator. Enter our AI model.

For example. p.q works as a trap-door if both p and q have the same
length. That is not possible, not even to be close (well-known
vulnerability).

Conversely, if (say) p > q, then the leftmost binary digits of p are

totally exposed, verbatim. There is no trap-door for that part.

Thus, RSA falls to algebra + AI. Pure obfuscation, this work reveals
-- not an impossible problem. The quantum Eq.(4) is not needed, as
values in the set N reflect them. So, prime values can be found non
sequentially.

2.8. Rigorous AI classifications

All the foregoing is inherently an algebraic-numeric calculation in the

set N (negative values are accounted for by encoding) which (in a
finite set) is well-known to be isomorphic to set Q (Section 4.2). This
classification is essentially computable by humans or machines.

The sets N and Q (being isomorphic in finite sets within AL), become
rigorous classifications in the AI model, corresponding to
intersubjective interpretations by humans.

2.9. Arbitrary-length rational values

19
All observations in nature are arbitrary-length rational values, as is
well-known.

The AL integers calculated by digital QC, therefore, are able to

harmonize all observations found or made in nature exactly, as a zero
error model. Any error may be due to different intersubjective
interpretations by humans, linking to correspondence differences in
the AI model, or accounted for as a stimulation out of range.

Everything else can be produced from such AL integers as a

simulation, as an interpolation, such as complex numbers, irrational
numbers, real-numbers, or any other number. All numeric calculations
can be done exactly using AL integers, and only “translated” at a
final step, to reduce error.

2.10. How to do quantum computation with qubits

The solution described here is to use the end configuration of qubit

patterns, as an eigenvalue.

The eigenvalues represent prime numbers, when using the (prime

number) Potential V(M, x) in the Equation. This is proven in [25],
using a tri-diagonal representation for the operator H, at any AL
discretization, as cited above.

This looks incredibly powerful as noted in the Introduction, as it

operates with qubits, and also incredibly difficult. However, it is
given by the eigenvalues -- as generalized resonances including
superposition and entanglement -- which final patterns are exact and
measurable, in an algebraic-numeric computation using only the set
B={0,1}.

Mathematics will feel the pressure to change, as values

now must be discrete and never continuous [26].

3. Results
20
We introduced digital QC, as an algebraic-numerical algorithm. The
algorithm presents the following results:

3.1. The PFT, by Gauss in 1873, states that every integer greater than
1 can be uniquely represented as a product of integer prime numbers,
which is irrespective of integer length and factor order.

3.2. Digital QC shows that prime numbers are eigenvalues of the set
N.

3.3. Digital QC shows that prime numbers are not necessarily

sequentially calculated.

3.4. The algebra of integer arithmetic reflects the quantum properties

of values found by digital QC.

Thus:

3.6. With algebra, given any RSA modulus, we can calculate the two
prime numbers that compose it, irrespective of integer length
(obfuscation).

3.7. Ofuscation is not important to break a RSA modulus.

3.8. Digital QC does not have to be used, nor AI. Algebra of integer
arithmetic is all one needs.

This cannot trigger any panic [18l], as it particularly used a RSA

modulus reserved for tests and did not use any private information.

4. Discussion

An AL integer is physically observable -- just use a computer and

compute any values using the set B={0,1} or any base such as
decimals. Even by hand. Humans can use digital QC!

We expect a new path for digital QC + AI. To clarify, that is not a

"feeling" but is based on AI itself.

21
First, what does digital QC solve? It is based on exact relationships
given by interconnecting AL integers.

This leads to diophantine equations to solve exactly the Schrödinger

equation by an eigenvalue/eigenvector problem in all cases,
time-dependent or not. An eigenvalue is the only observable solution.
It can also be its unique value.

Thus, prime values are eigenvalues of the potential of the Equation as

V(M, x), as defined in linear algebra, and easily calculated using a
disclosed matrix tridiagonal form of the Equation with AL integers.
The digital QC solves the apparently impossible problem of
calculating using qubits. That makes all AL integers accessible.
Meanwhile, it makes irrational values, mathematical real-numbers,
imaginary numbers, and any other value to be outside an AL finite
integer of length M.

Hence, all values that solve the Equation must belong to the set N,
defining diophantine algorithms as the only possible solutions.

AI is then useful to search in such a large space for the values that
numerically satisfy a particular diophantine algorithm.

This process does not have to use any human output, as AlphaGo
shows. This will be published elsewhere, as particular numerical
realizations of a general algebraic rule.

We conclude that obfuscation has been used to “protect” a RSA

modulus, but a computer can use an AI model (as Section 2.5
presents) and digital QC to effectively search and find a large prime
number that humans cannot find and has to be sequential in classical
computers —- so classical computers cannot find either.

Digital QC can produce approximate mathematical real-numbers,

irrational values, or imagined complex values as simulations -- e.g.,
with a coprocessor.

22
In the future, the market can produce directly, thus, what is physically
observable -- with digital QC and avoid analog QC, cryogenics,
imaginary numbers, special materials, and so on.

Avoiding simulation to avoid computational errors ... and slower and

more expensive computation of values.

We reach "wheels within wheels" -- everything synchronized, which

allows different values to interconnect perfectly.

There has been an unfortunate tendency in mathematics to view

numbers as “names”.

However, digital QC and QM invite us to consider a rigorous model

of nature, which cannot admit error. Thus, a number can still be
considered free creations of the human mind; albeit they serve as a
means of apprehending more easily and more sharply the difference
of things, as a value. This leads us to view a number as a 1:1 mapping
between a name and a value, resolving any apparent conflict.

Digital QC may benefit from a mechanical analogy. It is a sort of

"Antikythera mechanism" -- which is a well-known 2,000 years old
mechanical scientific calculator for astronomy. Or, with behavior like
"LEGO" blocks, where every value fits wherever one wants -- "click
mathematics".

Creativity can hopefully increase, and this will drive innovation,

using intuition.

Digital QC is not a 0 or a 1 decision at every step until the ... last step.
This can be optimized in hardware for digital QC, with only one
switch to create a 2-level system (1 or Z). The chips already exist,
they are called tri-state --and one does not use the 0 state ... until the
very end.

This work describes digital QC, working today, reaching 10991 states
and more, delivering AL prime numbers, with non physical

23
limitations and with no temperature dependance. The solution has the
steps documented in Section 2.

All physical models using digital QC can be cast as an eigenset

algorithm Hψ=Eψ, solving the Equation and providing a way to
measure qubit patterns observably, using eigenvalues, which can be
solved comprehensively as a diophantine algorithm in AL integers.

The Schrödinger Equation is, thus, included as the basis of digital QC.

Digital QC shows that prime numbers are the eigenvalues of the set
N. Thus, a formula for prime numbers has been calculated.

This was searched in this work in order to predict prime values,

which has not been successful so far in pure Mathematics (see
Introduction).

Physics, then, by adding a connection to Mathematics, using the

Schrödinger equation, can link the Equation potential with its
eigenvalues.

The Equation potential is in Physics, the eigenvalues are in

Mathematics.

So, Mathematics has to follow Physics, if it wants to follow nature.

That is the foreseeable end of "imagined Mathematics"; we would
have to recycle many textbooks and reform education, training the
workforce to use AL integers exclusively.

QM shows itself (with digital QC) to be our best model of

representing ontic reality, which is done using AL integers.
Continuity cannot exist in pure Mathematics, or be supported as an
alternative. QM and QC affirm that there is no continuity. There must
be gaps between values, and -- therefore -- continuity does not exist in
nature, which can be described by discrete values.

24
No rejection is possible, no falsification of what our QM model is —-
that would be a medical illness (i.e., psychosis).

As technology evolves and becomes technologically-agnostic,

no-target cryptography (NTC) should in the future not allow threats as
considered here, in spite of the evolution of technology.

We believe that our AI model should have an increasing experimental

importance. For example, with the QC powerful insight that one is
looking for eigenvalues in the calculations, Eq.(4) is enabled,
non-sequential prime number search is possible, and the AI model
may break RSA even without using digital QC explicitly. As another
example, the inverse of a complex rotation matrix is always
approximate though, not so using AL integer numbers in our AI
model.

FUNDING
This research received external funding from Network Manifold
Associates, Inc. (NMA); Safevote, Inc; and PlanaltoResearch;
including the generous release of previous material nonpublic
information created by the author 42 years ago. The author declares
no conflict of interest.

ACKNOWLEDGMENTS
The author is indebted to the work of authors behind all the
references, late Mādhava of Sangamagrāma 250 years before Isaac
Newton, late Dr. d'Oliveira, late Cel. Dr. Amarante, Ann Gerck, at
PlanaltoResearch, Don Sawtelle of PlanaltoResearch in C
programming with finite but AL integers, and six anonymous
reviewers. ResearchGate and LinkedIn discussions and private
messages were also used in various threads, for “live” feedback,
important due to the physical isolation caused by COVID.

REFERENCES

25
[1] Gerck, E. “Algorithms for Quantum Computation: Derivatives of
Discontinuous Functions.” Mathematics 2023, 11, 68. also at
https://doi.org/10.3390/math1101006, accessed October 12, 2023.

[2] Vandersypen, L. M. K. et al. Experimental realization of

Shor’s quantum factoring algorithm using nuclear magnetic resonance.
Nature 414, 883–887 (2001).

[3] Martín-López, E. et al. Experimental realization of Shor’s quantum

factoring algorithm using qubit recycling. Nat. Photon. 6, 773–776 (2012).

[4] Bocharov, A., Roetteler, M., & Svore, K. M. Factoring with qutrits:
Shor’s algorithm on ternary and metaplectic quantum architectures. Phys.
Rev. A 96, 012306
(2017).

[5] Smolin, J. A., Smith, G. & Vargo, A. Oversimplifying

quantum factoring. Nature 499, 163–165 (2013).

[6] Dattani, Nikesh S., & Bryans, Nathaniel. Quantum factorization of 56153
with only 4 qubits. arXiv preprint
arXiv:1411.6758v3 (2014).

[7] Yan, S. Y. Quantum Algorithms for Integer Factorization. In: Quantum

Computational Number Theory. Springer, Cham 59–119 (2015).

[8] Peng, X. et al. Quantum Adiabatic Algorithm for Factorization and Its
Experimental Implementation. Phys.
Rev. Lett. 101, 220405 (2008).

[9] Burges, C. J. C. Factoring as Optimization. Microsoft Research

MSR-TR-200 (2002).

[10] Wang, B., Yang, X. & Zhang, D. Research on Quantum Annealing

Integer Factorization Based on Different
Columns. Front. Phys. 10:914578 (2022).

[11] Dridi, R. & Alghassi, H. Prime Factorization using Quantum Annealing

and Computational Algebraic Geometry.
Sci. Rep. 7, 43048 (2017).
26
[12] Xu, N. et al. Quantum Factorization of 143 on a Dipolar-
Coupling Nuclear Magnetic Resonance System. Phys.
Rev. Lett. 108, 130501 (2012).

[13] Li, Zhaokai et al. High-fidelity adiabatic quantum computation using

the intrinsic Hamiltonian of a spin system:
Application to the experimental factorization of 291311.
arXiv preprint arXiv:1706.08061 (2017).

[14] Xu, K. et al. Experimental Adiabatic Quantum Factorization under

Ambient Conditions Based on a Solid-State
Single Spin System. Phys. Rev. Lett. 118, 130504 (2017).

[15] Dhaulakhandi, R., Bikashi, K. B., Seo, F. J. Factorization of large tetra

and penta prime numbers on IBM quantum processor. Published in
arxiv.2304.04999. 2023.

[16] Gerck, E; Brito Cruz, C. H.; “Eigenvalue translation method for mode
calculations”. Published at Applied Optics 18(9):1341, May 1979.

[17] Aaronson, S. “What Makes Quantum Computing So Hard to

Explain?”, published in
https://www.quantamagazine.org/why-is-quantum-computing-so-hard
-to-explain-20210608/, accessed on May 22, 2023.

[18] Herman E. and Butler A., “Prosperity at Risk: The Quantum Computer
Threat to the US Financial System”. Hudson Institute Apr 3, 2023, published
at
https://www.hudson.org/technology/prosperity-risk-quantum-computer-threa
t-us-financial-system . Accessed January 18, 2024.

[19] Gisin, N; “Indeterminism in Physics, Classical Chaos and Bohmian

Mechanics: Are Real Numbers Really Real?.” Springer Nature, Erkenn
86, 1469–1481. https://doi.org/10.1007/s10670-019-00165-8, 2021, accessed
on March/2/2024.

[20] Gerck, E.; Gallas, J. A.; d'Oliveira. A. B.; “Solution of the Schrödinger
equation for bound states in closed form”. Published in Physical review A,

27
Atomic, molecular, and optical physics, 26:1(1). DOI:
10.1103/PhysRevA.26.662, accessed on May 13, 2024.

[21] E. Gerck, A. Gerck, “QC Algorithms: Faster Calculation of Prime

Numbers”, published at
https://www.researchgate.net/publication/373516233/ . Accessed January 12,
2024.

[22] 3.3: The Schrödinger Equation is an Eigenvalue Problem - Chemistry

LibreTexts, published at
https://chem.libretexts.org/Bookshelves/Physical_and_Theoretical_Chemistr
y_Textbook_Maps/Physical_Chemistry_(LibreTexts)/03%3A_The_Schrodin
ger_Equation_and_a_Particle_in_a_Box/3.03%3A_The_Schrodinger_Equati
on_is_an_Eigevalue_Problem, accessed May 13, 2024.

[23]
https://www.geeksforgeeks.org/how-to-generate-large-prime-numbers-for-rsa
-algorithm/, accessed on March/2/2024. Every prime number can be directly
used as C in Eq.(4).

[24] https://github.com/HigherOrderCO/Bend, accessed on May 13, 2024.

[25] Gerck, E.; d'Oliveira, A. B.; “Continued fraction calculation of the

eigenvalues of tridiagonal matrices arising from the Schroedinger equation”.
Published in Journal of Computational and Applied Mathematics, Volume 6,
Issue 1, March 1980, Pages 81-82. Also at DOI:
10.1016/0771-050X(80)90020-0, accessed on May 13, 2024.

[26] Cepelewicz, J., “How Failure Has Made Mathematics Stronger”.

Published in
https://www.quantamagazine.org/how-failure-has-made-mathematics-stronge
r-20240522/, accessed on May 23, 2024.

28