Quick read · plain language
A Better Way to Invest in New Kinds of Computers
A plain-language version of “The Problem-Substrate Mapping: A Framework for Honest Computational Investment”
The short version
This paper says we should stop putting most of our money into one type of quantum computer. A quantum computer uses the strange rules of tiny particles to do math. It matches different computing jobs to the physical machines that naturally do them best. It then gives five questions to test whether any new computing technology is worth funding.
Why the old way of investing is wrong
The paper says the quantum computing industry has spent 35 billion dollars. It has not produced any machine that can be sold for real work. The authors wrote three earlier papers about this failure. This paper asks a new question. What should we actually build? What should investors pay for? The answer is not one technology. It is a mix of many technologies. Each technology should solve the kind of problem that its own natural physics handles best. The authors call this the Problem-Substrate Mapping.
Six kinds of computing jobs
The paper lists six types of problems that people pay computers to solve.
Finding the best answer among many choices is called optimization. This job matches machines that naturally settle into low-energy states, like Ising machines. Ising machines are devices made of connected parts that find the lowest-energy arrangement, like a compass needle finding north.
Multiplying big tables of numbers is called linear algebra. This job matches machines that use light to do calculations, called optical processors. Optical processors use light beams instead of electric current.
Working with uncertainty and chance is called probabilistic inference. This job matches brain-like chips called neuromorphic chips. Neuromorphic chips work like a brain, using many simple units that send brief signals. This job also matches tiny random devices called p-bits. P-bits are tiny switches that flip between zero and one with adjustable odds, like a weighted coin.
Copying the behavior of tiny quantum systems is called quantum simulation. This job matches special devices called quantum simulators. Quantum simulators copy how tiny particles act. They are not general quantum computers. This is the only place where quantum physics truly helps.
Breaking codes is called cryptography. This job might need full quantum computers with error correction, called error-corrected quantum computers. These fix mistakes using extra parts. They do not exist yet and will not for many years.
Running everyday software is called general-purpose computing. This job matches improved ordinary chips. It might also match reversible chips far in the future. Reversible chips recover energy instead of losing it as heat, like pedaling a bike backward to get your effort back.
What money should go where now
The paper gives a plan for where to put research money. In the next one to three years, about a quarter should go to thermodynamic solvers for optimization. Thermodynamic solvers use natural heat flow to find answers. About a fifth should go to optical processors and to neuromorphic chips each. Some smaller amounts should go to quantum simulators, p-bit networks, and better ordinary chips.
In three to seven years, quantum simulators should get the biggest share. In seven to fifteen years, reversible computing and room-temperature quantum devices should lead.
What the five questions are
The paper gives five questions to test any new computing technology. What problem does it solve? What natural physics does it use? How much energy does each answer cost? What ordinary machine does it compete with? What exact result will it show by what date?
The paper also lists warning signs. Be careful if someone claims quantum advantage without saying the exact problem and energy cost. Be careful if they use fake test problems with no business value. Be careful if they compare against old ordinary methods instead of the best ones. Be careful if they keep saying "five more years" every year. Be careful if they hide their hardware from outside checkers.
Good signs include giving exact energy numbers, letting outsiders verify, admitting the true costs of fixing errors, comparing against the best specialized ordinary hardware, and meeting deadlines on time.
What this paper does not prove
The paper does not prove that any of these technologies will succeed. It says the thermodynamic analysis shows some paths are possible in principle but far from commercial use. Thermodynamic analysis looks at heat and energy limits in a system. It uses a math tool called a Hamiltonian. A Hamiltonian is a formula for all the pushes and pulls on moving parts in a system. It does not claim the portfolio will produce revolutionary industries within five years. It says honest returns measured in energy per answer matter more than exciting stories. It leaves open whether private investors or only government agencies should fund the longest-term research. It also leaves open whether any quantum approach to cryptography will ever be worth its enormous energy cost.
Words used here
- Hamiltonian
- a formula for all the pushes and pulls on moving parts in a system
- Ising machine
- a device made of connected parts that find the lowest-energy arrangement, like a compass needle finding north
- Joules-per-solution
- the amount of energy a machine uses to finish one complete computing job, like the gas a car burns for one trip
- Neuromorphic
- built to work like a brain, using many simple units that send brief signals to each other
- Optical processor
- a computer that uses light beams instead of electric current to do calculations
- p-bit
- a tiny random switch that flips between zero and one with tunable odds, like a coin whose weight you can adjust
- Reversible computing
- a way of running logic gates so the energy used to move information can be recovered instead of lost as heat, like pedaling a bike backward to get your effort back
- Substrate
- the physical material or system that does the actual computing work
Learn more in plain language
- Performance per wattWikipedia (English, more technical)
- Silicon photonicsWikipedia (English, more technical)
- Bio-inspired computingWikipedia (English, more technical)
- Quantum simulatorWikipedia (English, more technical)
- Hamiltonian mechanicsSimple English Wikipedia
- Ising machineWikipedia (English, more technical)
This plain-language version was written by an AI model from the text of the paper and checked against plain-language rules: everyday words, no mathematics, no unexplained names. It leaves details out and can still be wrong; the paper is the source of record. To report a problem, open the paper, select the passage and choose Report.
For specialists: the technical digest, quoted from the paper
In the paper's words
Phase IV of the Qubit Delusion series. Translates the critique into a concrete investment portfolio: matching computational problem classes (optimization, linear algebra, probabilistic inference, quantum simulation, cryptography, general-purpose) to optimal physical substrates (Ising machines, optical processors, neuromorphic, analog quantum simulators, reversible classical, fault-tolerant QC). Proposes a standardized five-question evaluation framework with red flags and green flags.
The abstract, verbatim, and its opening sentence is the claim. Every line below is quoted from the text.
Numbers it reports
Near-term investment thesis (1-3 years): Deploy neuromorphic processors for edge AI --- always-on sensing, keyword spotting, anomaly detection --- where the energy budget is severely constrained (microwatts to milliwatts).
from 3.3 Neuromorphic and p-Bit ProcessorsStill --- a 10× to 100× improvement in the energy efficiency of general-purpose computation would be transformative.
from 3.5 Reversible Classical ComputingLong-term investment thesis (7-15 years): Fund fundamental research in reversible and adiabatic computing as the long-term path to energy-efficient general-purpose computation.
from 3.5 Reversible Classical ComputingGoogle's Willow processor (2024) demonstrated error correction below the surface-code threshold --- a genuine scientific achievement --- but at a scale (105 qubits) that is $10^{5}\times$ smaller than what is needed for useful computation.
from 3.6 Fault-Tolerant Quantum ComputingThe resulting portfolio allocates ~45% to thermodynamic/analog, 25% to optical, 20% to neuromorphic, and ~10% to quantum — a near-inversion of current R&D allocation where gate-model QC absorbs ~60%.
from AbstractComputational demands: For 2048-bit RSA, Shor's algorithm requires approximately 4,100 logical qubits and $10^{9}$ Toffoli gates.
from 2.5 Cryptography and Number Theory
How it is done
Shor's algorithm provides an exponential quantum speedup for these problems.
from 2.5 Cryptography and Number TheoryMemristive neuromorphic systems remain at the research prototype stage but have demonstrated proof-of-concept matrix multiplication at ~10 fJ per operation.
from 3.3 Neuromorphic and p-Bit ProcessorsInformation is never erased except at final readout, so the Landauer bound is paid only once per computation, not once per operation.
from 3.5 Reversible Classical ComputingThe joules-per-solution crossover for Shor's algorithm on 2048-bit RSA is estimated at $\sim 10^{7}$ physical qubits, requiring a cryogenic infrastructure of unprecedented scale.
from 3.6 Fault-Tolerant Quantum Computing
What would falsify it
Falsifiable milestones for each allocation.
from 1. Introduction: From Critique to PortfolioAny technology that cannot state a falsifiable joules-per-solution milestone: This is the acid test.
from 4.4 What Is NOT in the Portfolio(Not "conventional CMOS" --- the BEST classical alternative, including specialized hardware: FPGA, ASIC, reversible) What is the falsifiable milestone with a specific timeframe?
from 5.1 The Five Questions
Where the claim stops
Approximately 45% goes to thermodynamic and analog computing (Ising machines, p-bit networks, analog quantum simulators), 25% to optical/photonic computing, 20% to neuromorphic, and only 5-10% to fault-tolerant quantum --- a near- inversion of the current allocation, where gate-model quantum computing absorbs roughly 60% of advanced computing investment.
from 6. Conclusion: The Honest PortfolioThe honest portfolio does not promise exponential speedups or revolutionary new industries within five years.
from 6. Conclusion: The Honest PortfolioThis is the one problem class where quantum physics provides a genuine, in-principle exponential advantage --- and it does not require fault tolerance, error correction, or universal gate sets.
from 2.4 Quantum Simulation
Where to read
- 1. Introduction: From Critique to Portfolio281 words · 1 min
- 2. Problem Classes and Their Physical Signatures1,077 words · 5 min
- 3. The Substrate Portfolio1,232 words · 5 min
- 4. The Investment Portfolio534 words · 2 min
- 5. The Evaluation Framework284 words · 1 min
- 6. Conclusion: The Honest Portfolio384 words · 2 min
Read from the paper itself: 27 sections, 3,877 words. Nothing in this digest is written by a model; a line that is not a sentence of the paper is not shown. Have a specific question? Ask this paper.
The full text · Ask the whole corpus
Keep exploring
One body of research, three doors.