IBM's quantum computer simulating Möbius molecules is fake.

Home page
Quantum computer is useless

IBM's quantum computer (= still Not a computer ) could only give random useless numbers with No ability to compute anything nor clarify molecular behavior, contrary to overhyped fake news.

(Fig.1)  Quantum computers cannot even factor 21, much less calculate molecules, contrary to fake news.

IBM, Oxford's fake simulation of Möbius molecule.

Today's error-prone useless quantum computers cannot even factor the simple 21, much less calculate ( Möbius ) molecular energy, contrary to fake news.

The recent overhyped fake news in 2026 falsely claims that IBM or Oxford's quantum computer could simulate and clarify some new small Möbius-like molecule C₁₃Cl₁₂, which may be hard for a classical computer.

But today's error-prone useless quantum computers can Not even factor the simplest 21, so it is much more impossible to simulate or calculate molecular energy or properties, contrary to the above fake news.

IBM's useless 72 qubits are Not a computer.

In this research, IBM used only 72 qubits (= one qubit can take only 0 or 1 value ), which is far from a practical quamtum computer that needs millions of qubits.

Furthermore, IBM used only 72 qubits (= one qubit can take only 0 or 1 value,  this-lower ), which is still Not a quantum computer, and far from millions of qubits required for a practical quantum computer ( this-8th-last-paragraph ).

This-5~6th paragraphs say  -- Only useless 72 qubits
"the team relied on a quantum computer using 72 qubits."
"These may Not seem like viral, headline-making studies"

Quantum mechanics cannot predict molecules.

Today's unrealistic quantum mechanics cannot solve its Schrödinger equations nor predict any multi-electron atomic or molecular energies, so useless, whether (fictional) quantum or classical computers are used.

First of all, the present mainstream physics = unrealistic quantum mechanics is completely useless, can Not predict any energies nor properties of molecules.

Quantum mechanics can Not solve any Schrödinger equations nor predict energies of any multi-electron molecules ( this-p.7-upper,  this-p.2-last-paragraph )

Quantum mechanics can only choose fake trial ( basis set ) wavefunctions for unsolvable Schrödinger equations, which cannot predict any molecular behavior.

Quantum mechanics has to artificially choose fake trial (= basis-set ) wavefunctions with many free parameters out or infinite choices ( this-4~5th-paragraphs ) to find ones giving the lowest atomic energy (= Not true wavefunctions ), which takes infinite time ( this-p.13-last,  this-p.5-last ).

This mainstream quantum mechanical approximate method just choosing fake wavefunctions and adjusting free parameters is called variational method.

↑ Guessing or choosing fake trial wavefunctions or solutions for unsolvable Schrödinger equations means quantum mechanics has No ability to predict any atomic nor molecular energies ( this-p.6,  this-p.3-lower,  this-p.1-last ).

Impractical quantum approximation = CI.

This IBM research used the impractical quantum mechanical approximation called configuration interaction (= CI ) which just chooses fake wavefunctions that cannot predict anything.

This IBM research used very time-consuming impractical quantum mechanical approximation (= for unsolvable Schrodinger equations that cannot predict anything ) called (selected) configuration interaction (= CI ), as shown in this IBM paper-p.35-3.5-Quantum calculations called sqDRIFT used the ( rough approximate )selected configuration interaction.

↑ In this approximate quantum CI, they artificially choose fake trial wavefunctions consisting of multiple (unphysical) antisymmetric wavefunctions expressed as Slater determinants ( this-p.5-(9)(10),  this-p.7-p.8 ).

Quantum computers give only random numbers.

This IBM error-prone quantum computer could only generate random useless numbers called sampling as artificial fake wavefunctions' parameters without calculating nor predicting molecular energy.

↑ This research's IBM (fake) quantum computers was used only to randomly choose (= generate ) these free parameters of artificially-chosen wavefunctions or Slater determinants (= called sampling ), which is useless, cannot predict any molecular energies nor behavior.

This-IBM paper-p.37-2nd-paragraph, p.38-1st-paragraph say
"which randomizes the Hamiltonian terms to be implemented within the quantum circuit"
"SqDRIFT leverages the quantum processor solely as a sampling machine, and the subsequent diagonalization is executed classically (= actual molecular energy calculations were performed by a classical computer )"

After randomly choosing these Slater determinants (= fake trial wavefunctions with many different parameters ), this CI method linearly connects these chosen determinants by multiple coefficient (= C, a, b ) free parameters.

A classical computers was needed to optimize free coefficient parameters connecting randomly-chosen fake trial wavefunctions to calculate and get the lowest-energy molecular states (= by diagonalization ) instead of useless quantum computers.

↑ The coefficient free parameters (= c, a ) connecting fake trial wavefunctions or determinants (= randomly chosen by the useless quantum computer ) were optimized to give the lowest molecular energy within the chosen trial wavefunctions or determinants by variational methods called diagonalization ( this-(61),  this-(7.10),   this-p.15-(3.58),  this-p.5-(2.4) ) by an ordinary classical computer instead of today's useless quantum computers.

↑ This most important, time-consuming calculation of molecular energies, optimizing (free) parameters to find the lowest molecular ( Hamiltonian ) energy (= called diagonalization, this-p.7-p.8 ) was conducted by an ordinary classical computer ( this-3.  this-p.38-1st-paragraph,  this-p.3-left-upper,p.5 ).

This-p.4-1st-paragraph says  -- Useless quantum computer
"Thus, the quantum part of QSCI lies solely in the ( random ) selection of determinants (= trial wavefunctions ) for the SCI expansion... since the quantum device is only used for ( random ) selection, while energies (= the most important calculations ) are obtained classically (= a classical computer )"

Because today's error-prone useless quantum computers cannot even factor the simplest 21, much less calculate molecular energies.

No quantum computer's advantage

The useless quantum computer could only repeat generating or sampling random parameters of trial functions, until they luckily got the good molecular energy (= whose calculations and parameter-optimization must be done by a classical computer ).

The more different trial wavefunctions (= or determinants ) were randomly generated or chosen (= by the impractical quantum computer ), the more likely they could find and artificially choose the lower (= more correct ) molecular energy ( this-p.4-Algorithm1-last ), because they could optimize more free coefficient parameters (= by a classical computer ) connecting more trial wavefunctions or determinants.

After they randomly collected some numbers of (fake) trial wavefunctions or determinants (= parameters expressed by random qubits' numbers ) by the IBM quantum computer, this could get almost the same molecular energy value as ones obtained by trial wavefunctions chosen by a classical computer.

So there is No quantum computer advantage nor utility, and the quantum mechanics cannot predict any molecular energy, whether quantum or classical computers are used.

Quantum computers cannot predict energy.

This IBM research paper did Not say their (impractical) quantum computer could predict molecular energy nor outperform the classical computer.

This-lower quantum computing meets chemistry says
"The quantum-assisted results (= actual calculation must be done by a classical computer though ) agreed with (= Not outperform ) classical calculations"
"The team suggests that quantum hardware may (= uncertain word, so still useless ) become increasingly useful for modeling system"

This IBM research paper did Not say "quantum computers could predict some molecular energies" nor "quantum computers could calculate molecule energies faster or more correctly than a classical computer".

This-13~14th-paragraph says  -- No quantum advantage
"They do say that they surpassed what exact classical simulation can do, but that isn't the relevant question, Aaronson explained"  ← No clear mention of quantum computer's advantage.

Unreal quantum mechanical molecular model.

The unphysical quantum mechanical artificially-chosen fake wavefunctions or unreal quasiparticle model can Not predict nor clarify any molecular mechanisms, contrary to fake news,

These artificially-chosen fake trial wavefunctions or Slater determinants are unphysical, unable to clarify any real molecular mechanisms, contrary to the overhyped fake news.

Actually, this IBM research also tried to express the molecule observed by atomic force microscope (= AFM ) by a fictional quasiparticle ( this-abstract ) or one-pseudo-electron DFT model with artificially-chosen exchange-pseudo-potential, which also cannot predict any molecular energy nor clarify real atomic properties.

Atomic force microscopes are deadend.

Even this latest IBM research could observe only a small flat molecule by the atomic force microscope with only one useless probe tip with No progress for 40 years.

Atomic force microscope technology has been deadend with No progress, stuck in only one useless probe tip for 40 years due to these unphysical impractical quantum mechanical models lacking real atomic shape.

↑ Actually even this latest research could observe only small flat molecule by a atomic force microscope with only-one probe tip ( this-p.30~ = No progress,  this-lower-Challenges ).

Because they intentionally try Not to develop useful multi-probe atomic force microscopes that can manipulate any 3-dimensional molecules, only to protect the old impractical quantum mechanics exploited for investment fraud easily collecting money just by spreading overhyped fake news.

Fake parallel-world quantum advantage

Today's fake useless quantum supremacy or advantage is based on false idea that quantum computers can get random (meaningless) numbers through many (unseen) parallel-world qubits' states that may be hard for a classical computer to emulate.

The fake quantum computer's advantage in this research originates from the unfounded claim that IBM's 72-qubit quantum computer could get random (useless) numbers ( this-2nd-paragraph ) via (fictional) quantum superposition of 2⁷² parallel-world qubits' states that may be hard for a classical computer's single world to emulate.

This or this-3rd-last and 2nd-last paragraphs say  -- Unreal parallel worlds
" IBM's quantum computer represents the states using quantum bits, called qubits, which can represent a superposition of different quantum possibilities (= unreal parallel worlds ), it can perform bigger calculations at less cost (= false )."

"So the team was able to scale up the calculation and confirm that, after a certain point, the resulting electron clouds looked more or less the same (= quantum and classical computers happened to give the same results with No quantum advantage,  this-lower-quantum computing meets chemistry )"

↑ As seen in Google's fake quantum supremacy giving only almost random meaningless numbers including a lot of errors, today's error-prone quantum computers can only give random meaningless numbers with No ability to calculate molecular energies nor factor numbers.

A classical computer can get random numbers more quickly than a quantum computer, so No quantum advantage.

This IBM's useless 72-qubit quantum computer just generated random numbers, and actual molecular energy calculation had to be performed by a classical computer with No quantum advantage.

↑ This IBM research paper ( this-last-link ), this ↓

p.1-abstract mentions  -- No quantum advantage
Unreal quasiparticle model, and said nothing about a quantum computer's advantage over a classical computer.

p.33-3.4 says  -- Unreal quantum model
One-fake-electron DFT model with artificially-chosen empirical exchange-correlation energy functional (= fake potential ) called ωB97X-D that could just qualitatively (= Not quantitatively ) reproduce some molecular states, which cannot predict any true molecular energies.

p.35-3.5-Quantum calculations say  -- Impractical quantum CI
"The recently-proposed SqDRIFT quantum algorithm (36) is a quantum-computing counterpart of classical (impractical) selected Configuration Interaction (= CI ) approaches"

p.37-1st-paragraph says  -- Impractical quantum computer
"To execute SqDRIFT on a quantum processor, the molecular Hamiltonian had to be mapped onto the quantum hardware... which resulted in 72-qubit circuits.... would be impractical on current quantum hardware."

p.37-2nd-paragraph says  -- Useless random sampling
"To sidestep these limitations, we employed the qDRIFT protocol, which randomizes the Hamiltonian terms to be implemented within the quantum circuit,"

p.38-1st-paragraph says  -- No quantum computer calculation
"As described above, SqDRIFT leverages the quantum processor solely as a ( random useless ) sampling machine, and the subsequent diagonalization (= calculate and find the lowest molecular energy ) is executed classically (= a classical computer was needed to calculate molecular energy = Hamiltonian energy after all )"

p.39-last-paragraph says  -- No quantum advantage
"The Dyon orbital obtained by SqDRIFT agrees well with (= did Not outperform ) the the classical calculations using CASSCF"

Quantum mechanics cannot get right energies

This research tries to rigorously calculate energies in only 12 (~ 33 ) electrons in all 112 electrons of C₁₃Cl₂, which cannot predict right total energy of the whole molecule, so meaningless research.

Quantum mechanics just artificially choosing fake trial wavefunctions cannot predict any molecular energies.

Furthermore, this research tries to choose rigorous (fake) trial wavefunctions (= by the impractical full-CI ) of only 12 electrons of all 112 electrons of the whole C₁₃Cl₂ molecule in CASSCF, which can not predict nor clarify the exact molecular energy nor properties.

↑ This-same paper-p.36-last-paragraph says  -- Too few electrons
"we performed SqDRIFT calculations on an IBM quantum processor carried out on an active space of 32 electrons in 36 orbitals (= out of the total 112 electrons ) for the neutral molecular system, and 33 electrons in 36 orbitals for the anionic (= unreal quasiparticle with fractional charges ) molecular system. Such an active space includes all carbon π electrons and per chlorine four electrons (= ignoring inner electrons ) that mix with the π-system of C₁₃C_l2. The molecular orbitals were obtained from a smaller-scale CASSCF calculation, based on an active space including 12 electrons in 12 orbitals"

↑ The actual time-consuming molecular energy calculation must be conducted by a classical computer (= so IBM's useless quantum computer generating only random number parameters for 32 electrons is meaningless, No quantum advantage ), so this research used (impractical) full-configuration-interaction (= full CI ) with artificially-chosen complicated trial wavefunctions in only 12 electrons (= too rough approximation ! ) out of the total 112 electrons of the entire C₁₃C_l2 molecule.

So this research cannot get the exact total energy of the entire molecule, because the calculated total energy is artificially changed by how many electrons are chosen as the active space where full-CI method is applied ( this-p.16,  this-(7.150),  this-3rd-paragraph ).

No quantum computer advantage

↑ So No quantum computer advantage over a classical computer (= random numbers generated by a quantum computer had No guarantee of giving better molecular energy ), and this unreal quantum mechanical methods cannot predict any molecular energy nor clarify real molecular mechanisms.

This-4th-last-paragraph says  -- Useless quantum computer
"At this stage, SqDRIFT (= useless quantum computers just randomly sampling with No ability to calculate energies, error-prone, this-last-paragraph ) does Not aim to replace classical methods;... And thanks to its better scaling behavior, will (= uncertain future, so still unrealized ) soon surpass some of the abilities of its counterparts"

Classical computer better

In another recent IBM research paper in 2026 using the same Heron quantum computer as this, a classical computer choosing better trial (= basis-set ) wavefunctions could get better (= lowest ) molecular energy than the quantum computer's randomly chosen wavefunctions (= this-p.12-left,p.13-Fig.6 showed classical computer's methods = FCIQMC got the lower better molecular energy than random quantum computer's sampling = (Ext-)SqDRIFT ).

Feel free to link to this site.