In This Module — 8 Chapters
Every classical computer ever built — from the room-sized ENIAC of 1945 to the Apple M4 of 2024 — does exactly one thing at its core: it processes bits, each either a 0 or a 1. These bits are physically represented by transistors, tiny electronic switches that are either open or closed. By 2024, the Apple M4 contains 28 billion of these switches in a chip smaller than a postage stamp, performing over a trillion operations per second. The entire digital world — every email, video, financial transaction, and AI model — is ultimately billions of these switches flipping states at incomprehensible speed. This architecture has driven the most extraordinary technological transformation in human history. But it carries within it the seeds of its own physical limit.
Key Insight
The Intel 4004 (1971) had 2,300 transistors. The Apple M4 (2024) has 28 billion — a 12-million-fold increase in a chip costing less per transistor by a factor of millions. No technology in history has improved at this rate for this long.
1945
ENIAC
17,468 vacuum tubes
28B
Apple M4
transistors (2024)
2nm
Modern transistor
10–15 silicon atoms
1. What is the fundamental unit of information that every classical computer processes?
AA byte
BA pixel
CA bit — a value of either 0 or 1
DA transistor state index
Correct — C. A bit is the most fundamental unit of classical computing — a value that can only ever be 0 or 1. A byte is simply 8 bits grouped together. Everything in the digital world is ultimately a sequence of these binary values being processed at speed.
2. What physical component represents a single bit inside a modern processor?
AA capacitor
BA transistor — an electronic on/off switch
CA silicon atom
DA logic gate array
Correct — B. A transistor is a semiconductor switch. When the gate voltage is on, current flows — representing 1. When off, no current flows — representing 0. Every bit in every processor is physically implemented this way.
3. Approximately how many transistors does the Apple M4 chip (2024) contain?
A42 million
B1 billion
C28 billion
D1 trillion
Correct — C. The Apple M4 contains 28 billion transistors in a chip smaller than a postage stamp. The Intel 4004 from 1971 had just 2,300 — making the M4 approximately 12 million times denser. This is the scale of Moore's Law compounded over 53 years.
Moore's Law
In 1965, Gordon Moore observed that the number of transistors on a chip was doubling roughly every two years at constant cost, and predicted this would continue. For fifty years, the entire semiconductor industry organised itself around delivering this prophecy — and it held. Twenty-five doublings amounts to a 33-million-fold improvement in computing power at roughly constant price. But by the mid-2010s, physics began to resist. Transistor dimensions had dropped below 10nm, where quantum effects begin to dominate. The doubling cadence slowed from 18–24 months to 3–5 years per node. Intel, TSMC, and Samsung continue advancing to 3nm and 2nm — but the era of reliable, predictable exponential doubling is over. Future performance gains now require architectural innovation, not simply smaller transistors.
Key Insight
Moore's Law was not a law of physics — it was a self-fulfilling engineering prophecy. It is ending not because engineers lack ingenuity, but because silicon transistors are approaching atomic dimensions where classical physics stops working reliably.
1. What does Moore's Law state?
AComputing speed doubles every year regardless of transistor count
BThe number of transistors on a chip doubles approximately every two years at constant cost
CChip prices halve every 18 months
DMemory capacity doubles every three years
Correct — B. Moore's 1965 observation was specifically about the density of components on a chip doubling approximately every two years at minimum cost. It was never a law of physics — it was an empirical observation that became a roadmap the industry funded and delivered for fifty years.
2. Why is Moore's Law slowing down today?
ASemiconductor companies have lost investment and research capacity
BConsumer demand for faster processors has declined
CTransistors are approaching atomic dimensions where quantum effects make reliable binary operation impossible
DSilicon has been replaced by inferior materials at smaller scales
Correct — C. At 2–3nm, transistors are only 10–15 silicon atoms wide. At this scale, quantum mechanical effects — particularly electron tunnelling — cause transistors to misbehave as binary switches. This is a fundamental physical limit, not an engineering or investment problem.
The Physics Wall
At nanometre scales, electrons stop behaving like classical particles and start exhibiting quantum behaviour. The most destructive of these is quantum tunnelling — electrons passing through insulating barriers they classically cannot overcome. When a transistor's insulating layer is only a few atoms thick, electrons tunnel through it even when the transistor is switched off, causing leakage that destroys reliable binary logic. IBM Research confirmed that off-state leakage increases by over 1,000x at 2nm compared to 28nm nodes. Alongside tunnelling, there is heat: every switching event dissipates energy, and packing 28 billion transistors into a fingernail-sized chip generates heat densities exceeding nuclear reactor fuel rods. These two problems — tunnelling and thermal density — are not engineering challenges that more investment can solve. They are consequences of how matter behaves at atomic scales.
Key Insight
Global data centres consumed approximately 240 terawatt-hours of electricity in 2022. The IEA projects this will exceed 1,000 TWh by 2026 as AI workloads grow. Microsoft and Google have already signed agreements to restart decommissioned nuclear plants to meet data centre demand.
1. What is quantum tunnelling in the context of transistors?
AA method used to speed up electron flow inside a chip
BA design technique to reduce heat in processors
CElectrons passing through insulating barriers they should be blocked by, causing leakage in off-state transistors
DA process for shrinking transistor gate lengths below 1nm
Correct — C. Quantum tunnelling occurs because electrons have wave-like properties. At dimensions below ~5nm, the insulating barrier is so thin that electrons tunnel through it even when the gate is off, making it impossible to maintain a clean 0 state. This directly undermines the binary logic that classical computing relies on.
2. Why does packing more transistors into a chip create a heat problem?
ALarger chips absorb more ambient heat from the environment
BEvery switching event dissipates energy as heat — billions of switches firing per second in a tiny area creates extreme thermal density
CTransistors generate heat only when in the on state
DSilicon becomes thermally unstable at nanometre dimensions
Correct — B. The Landauer principle states that erasing a bit requires minimum energy dissipation. Multiply this by 28 billion transistors switching billions of times per second, and the resulting heat flux per unit area exceeds that of nuclear reactor fuel rods — making cooling a fundamental constraint, not just an engineering challenge.
Quantum Mechanics
The quantum phenomena destroying classical computing are not only problems — they are computational resources. Wave-particle duality, confirmed experimentally from 1801 through to 1927, establishes that all matter has wave-like properties. This leads to superposition: a quantum particle genuinely exists in multiple states simultaneously until measured. A qubit exploits this, existing as a combination of 0 and 1 at the same time. Ten qubits can represent 1,024 states simultaneously; 300 qubits can represent more states than atoms in the observable universe. Then there is entanglement: when two qubits are entangled, measuring one instantly determines the other regardless of distance. The 2022 Nobel Prize in Physics was awarded specifically for experimental work conclusively confirming entanglement. These are experimentally verified phenomena that quantum computers harness as their primary computational resources.
Key Insight
Quantum computers do not work by "trying all answers at once." They work by engineering interference patterns — amplifying the probability of correct answers and cancelling out incorrect ones. Only specific problem classes with particular mathematical structure can exploit this approach.
1. What is quantum superposition?
AA qubit switching rapidly between 0 and 1 faster than classical bits
BA qubit storing both 0 and 1 in separate memory locations simultaneously
CA qubit genuinely existing in a combination of 0 and 1 simultaneously until the moment of measurement
DTwo qubits sharing the same physical location in a chip
Correct — C. Superposition is not a metaphor. A qubit in superposition is mathematically described as a combination of both states simultaneously. The probabilities only resolve to a definite value at the moment of measurement. This is one of the most rigorously tested results in all of physics.
2. What was the significance of the 2022 Nobel Prize in Physics for quantum computing?
AIt was awarded for building the first fault-tolerant quantum computer
BIt was awarded for experimental work conclusively confirming that quantum entanglement is real and cannot be explained by classical physics
CIt was awarded for developing the first quantum algorithm to beat a classical computer
DIt was awarded for discovering the qubit as a theoretical concept
Correct — B. Alain Aspect, John Clauser, and Anton Zeilinger shared the 2022 Nobel Prize for Bell test experiments that definitively closed all known loopholes and confirmed quantum entanglement beyond any classical explanation. Not theoretical, not contested — one of the most rigorously verified results in science.
3. How many simultaneous states can 10 qubits in superposition represent?
Correct — C. Each additional qubit doubles the number of states representable simultaneously: 2 to the power of 10 = 1,024. This exponential scaling is the source of quantum computing's potential advantage — 50 qubits can represent over a quadrillion states, and 300 qubits more states than atoms in the observable universe.
Unsolvable Problems
Beyond the physical wall, there is a mathematical one. Certain problem classes are not merely hard for classical computers — they are provably intractable. The Travelling Salesman Problem illustrates this: finding the optimal route through 50 cities requires evaluating a number of possible routes with 64 digits — more routes than atoms in the observable universe. This combinatorial explosion affects three critically important real-world domains: optimisation (logistics, drug discovery, portfolio management); quantum system simulation (molecular modelling, materials science — where simulating even a caffeine molecule requires more classical memory than exists on Earth); and cryptographic factoring (where Shor's quantum algorithm can break RSA encryption that classical computers cannot crack in the age of the universe). McKinsey estimates these three domains represent $450–850 billion in annual quantum computing value by 2035.
Key Insight
Richard Feynman (1982): "Nature isn't classical, dammit, and if you want to make a simulation of nature, you'd better make it quantum mechanical." Quantum computers can simulate molecules naturally because they are quantum systems themselves.
1. Why is the Travelling Salesman Problem intractable for classical computers at large scale?
AClassical computers cannot process geographic data efficiently
BThe problem requires too much memory to store city coordinates
CThe number of possible routes grows factorially — faster than any classical computer can search regardless of speed
DClassical computers cannot handle floating-point distance calculations at scale
Correct — C. For n cities, the number of possible routes is n factorial. At 50 cities, this number has 64 digits — more routes than atoms in the observable universe. No increase in classical computing speed changes this fundamental mathematical reality. This class of problem is called NP-hard.
2. Which of the following is NOT one of the three problem classes where quantum computing offers genuine advantage?
AOptimisation problems such as logistics and drug design
BSimulation of quantum systems such as molecules
CCryptographic factoring of large integers
DStreaming video and rendering graphics
Correct — D. Video rendering and graphics are well-suited to classical parallel computing via GPUs. Quantum advantage is specific to problems with exponential solution spaces (optimisation), quantum-mechanical structures (molecular simulation), and number-theoretic hardness (cryptography). Quantum computers will not replace classical computers for general tasks.
The Quantum Opportunity
Quantum computers are not faster classical computers. They are a fundamentally different kind of computer suited to a different class of problems — and making that distinction is the hallmark of genuine quantum literacy. Three algorithms define the primary domains of quantum advantage: Shor's Algorithm (1994) factors large integers in polynomial time, breaking RSA encryption that classical computers cannot crack; Grover's Algorithm (1996) searches unsorted databases in square-root time, delivering a quadratic speedup for search; and QAOA (2014) provides near-term approaches to combinatorial optimisation on current NISQ hardware. A quantum computation works not by trying all answers simultaneously but by engineering interference — quantum gates are arranged so that correct answers constructively amplify while incorrect answers destructively cancel, and measurement collapses the system to the correct result with high probability.
Key Insight
Quantum computers will operate alongside classical computers, not replace them. They handle the specific subset of problems where quantum mechanics provides genuine computational advantage. Knowing which problems fall into this category is the core skill of quantum business strategy.
1. What does Shor's Algorithm do that makes it significant for cybersecurity?
AIt detects intrusion attempts on quantum networks faster than classical firewalls
BIt encrypts data using quantum entanglement, making it unbreakable
CIt factors large integers in polynomial time, which would break RSA and most current internet encryption
DIt generates truly random cryptographic keys using quantum noise
Correct — C. RSA encryption is secure because factoring a 2048-bit number is computationally intractable for classical computers — it would take longer than the age of the universe. Shor's Algorithm solves factoring in polynomial time on a quantum computer, meaning a sufficiently powerful quantum machine could break RSA-2048 in hours.
2. How do quantum computers actually achieve correct results — what is the core mechanism?
ABy testing every possible answer simultaneously and selecting the best one
BBy using entangled qubits to communicate the answer between processors instantly
CBy engineering quantum interference so correct answers amplify and incorrect answers cancel out before measurement
DBy running the same calculation many times and averaging the results
Correct — C. Quantum computation works through interference. Quantum gates are designed so that the probability amplitudes of correct answers add constructively while those of incorrect answers add destructively. Measurement then collapses the system to the correct result with high probability. This is the fundamental mechanic — not brute-force parallelism.
Hardware Today
Quantum computing is not a theoretical future technology — it exists and is accessible now. IBM, Google, IonQ, Quantinuum, QuEra, and D-Wave all operate real quantum hardware, with IBM's Condor processor reaching 1,121 qubits and QuEra's Aquila operating 256 neutral atom qubits. The current era is described by physicist John Preskill's term NISQ — Noisy Intermediate-Scale Quantum: machines large enough to demonstrate genuine quantum effects, but with error rates too high for full-scale fault-tolerant computation. The landmark QuEra/Harvard paper (Nature, 2023) demonstrated 48 logical qubits encoded from 280 physical qubits — the most significant milestone toward fault-tolerant quantum computing achieved to date, cited over 500 times. IBM Quantum offers free cloud access: anyone can write a quantum program in Python using Qiskit and run it on real quantum hardware today.
Key Insight
McKinsey estimates quantum computing could generate $450–850 billion in annual value by 2035 across pharmaceuticals ($100–170B), finance ($60–80B), logistics ($40–70B), and materials science ($50–100B) — based on specific use cases with demonstrated theoretical quantum advantage.
1. What does NISQ stand for, and what does it describe?
ANew Integrated Silicon Quantum — the first silicon-based quantum chips
BNon-Isolated Sequential Qubit — a design approach for stable qubits
CNoisy Intermediate-Scale Quantum — current machines with genuine quantum capability but error rates too high for full fault-tolerant computation
DNear-Infinite Speed Quantum — the theoretical performance ceiling of quantum processors
Correct — C. NISQ was coined by John Preskill in 2018 to describe the current generation of quantum hardware: real, working machines with 50–1,000+ qubits, but with error rates that prevent running full-scale Shor's or Grover's algorithms at commercially useful scale. It is the current productive research era, not a failure state.
2. What was the significance of the QuEra / Harvard paper published in Nature in 2023?
AIt announced the first quantum computer to break RSA encryption
BIt demonstrated 48 logical qubits with below-threshold error rates — the most significant milestone toward fault-tolerant quantum computing achieved to date
CIt proved neutral atom quantum computers are definitively superior to all other qubit technologies
DIt was the first time a quantum computer solved an NP-hard problem faster than a classical supercomputer
Correct — B. The Bluvstein et al. paper demonstrated 48 logical qubits encoded from 280 physical qubits using neutral atom arrays, with error rates below the fault-tolerance threshold. Logical qubits — error-corrected qubits — are essential for practical quantum computing at scale. This was the first demonstration at this scale, making it the field's most important recent experimental milestone.
The Urgency
The most urgent quantum threat is not a future event — it is happening now. Nation-state adversaries are currently collecting and storing encrypted internet traffic. They cannot decrypt it today. When sufficiently powerful quantum computers arrive (current estimates: 2030–2040), they will decrypt everything collected. This is the harvest now, decrypt later strategy, and data with 10–20 year sensitivity — medical records, military intelligence, financial agreements — is already at risk. In response, NIST finalised three post-quantum cryptography standards in August 2024: CRYSTALS-Kyber (FIPS 203) for key encapsulation, CRYSTALS-Dilithium (FIPS 204) for digital signatures, and SPHINCS+ (FIPS 205) as a hash-based fallback. Organisations must audit their cryptographic implementations, identify long-sensitivity data, and begin migration planning now. The window for preparation is open — and it is not infinite.
Key Insight
NIST published FIPS 203, 204, and 205 in August 2024 — the first standardised post-quantum cryptography algorithms. Migration from classical to post-quantum cryptography is not a five-year problem. It is a present one. The UK NCSC, US CISA, and NATO have all issued guidance recommending organisations begin migration planning now.
1. What is the "harvest now, decrypt later" threat?
AA technique where quantum computers harvest energy from classical processors to power decryption
BA strategy for quantum computers to break encryption in real time as data is transmitted
CAdversaries collecting and storing encrypted data today so it can be decrypted by future quantum computers
DA data mining approach that extracts patterns from encrypted traffic without decrypting it
Correct — C. Nation-state actors with sufficient resources are storing encrypted internet traffic now, knowing they cannot decrypt it today. When fault-tolerant quantum computers capable of running Shor's Algorithm arrive, all of that stored data becomes readable retroactively. This makes data encrypted today vulnerable to future quantum attacks — the threat window is already open.
2. Which organisation finalised post-quantum cryptography standards in August 2024?
AIBM Quantum
BNIST — the US National Institute of Standards and Technology
CThe European Quantum Flagship programme
DThe UK National Quantum Strategy
Correct — B. NIST published FIPS 203 (CRYSTALS-Kyber), FIPS 204 (CRYSTALS-Dilithium), and FIPS 205 (SPHINCS+) in August 2024. These are the first standardised algorithms designed to resist attacks from quantum computers, and organisations are now expected to begin planning migration to these standards.
3. Why should organisations address post-quantum cryptography now, even though fault-tolerant quantum computers do not yet exist?
ABecause quantum computers already exist that can break current encryption
BBecause regulators currently fine organisations that use classical encryption
CBecause post-quantum algorithms are already faster and cheaper than classical ones
DBecause data with long-term sensitivity is already being harvested today, and cryptographic migration across complex systems takes years
Correct — D. Two reasons create urgency: first, the harvest now/decrypt later threat means today's data is already at risk from future quantum computers; second, cryptographic migration is not a quick fix — auditing systems, testing new algorithms, and updating infrastructure across an organisation can take 3–7 years. Waiting until quantum computers arrive is already too late for sensitive data.