Quantum Lab Session 5 Simulation Q13
🧲 Before we begin — what does a real qubit look like?
In simulations 1–12, we worked with perfect mathematical qubits — arrows on Bloch spheres, amplitude bars in circuits. Real qubits are very different. They are physical objects: superconducting circuits cooled to 15 millikelvin, individual atoms held in laser traps, photons bouncing between mirrors, or exotic quantum particles that don't even exist yet in useful form.

Building a quantum computer is not primarily a software problem. It is an extraordinary engineering challenge — keeping fragile quantum states alive long enough to compute with them, while also making millions of them work together reliably. The race between IBM, Google, IonQ, and Microsoft is fundamentally a race to solve this physical engineering problem.

Every qubit technology makes different trade-offs. There is no perfect qubit. Understanding these trade-offs explains why quantum hardware is still so hard, why error correction is necessary, and what the path to large-scale quantum computing actually looks like.
🌀 The engineering gap
Today's best quantum computers have ~1,000–4,000 physical qubits, but each has error rates of 0.1–1% per gate. To run Shor's algorithm on RSA-2048 securely, you'd need ~4 million physical qubits (to encode ~4,000 logical error-corrected qubits). The gap between where we are and where we need to be is enormous — and bridging it requires breakthrough hardware engineering across every qubit technology.
🧲 Quantum Hardware · Session 5 · Q13

The Qubit Zoo

Meet the real physical qubits that power today's quantum computers. Compare superconducting, trapped ion, photonic, and topological approaches — each with its own superpowers and limitations.

🧲 Superconducting
⚗️ Trapped Ion
💡 Photonic
📊 Comparison
🏆 Badge
🧲

Superconducting

Josephson junction circuits at 15mK. Used by IBM, Google. Fast gates (20–100ns), moderate coherence (100µs), scalable manufacturing.

⚗️

Trapped Ion

Individual atoms in laser traps. Used by IonQ, Honeywell. Long coherence (minutes!), slow gates (1–10µs), hard to scale to many qubits.

💡

Photonic

Qubits encoded in photons. Room temperature operation. Hard to make photons interact — limits two-qubit gates. Used by PsiQuantum.

🌀

Topological

Microsoft's bet on Majorana anyons. Theoretically immune to local noise. Still largely theoretical — first demonstrations only in 2023.

🧲
Wizzy · Quantum Guide
Superconducting qubits are the most advanced quantum computers today. IBM's Eagle, Osprey, and Condor processors — and Google's Sycamore — all use this technology. A qubit is a tiny superconducting circuit (a Josephson junction) cooled to 15 millikelvin. At that temperature, electrons pair up and flow without resistance, enabling quantum effects at circuit scale.
🌀 How cold is 15 millikelvin?
15 mK = 0.015 Kelvin = -273.135°C. The cosmic microwave background radiation (the coldest thing in the observable universe) is 2.7 K — 180× warmer than IBM's quantum computers. The extreme cold is needed to suppress thermal noise that would instantly destroy superposition at higher temperatures.

Superconducting Qubits — IBM & Google

Key Specs
🌡️ Temperature: 15 millikelvin
⏱️ Coherence (T2): 100–500 µs
⚡ Gate time: 20–100 ns
🎯 Gate fidelity: 99.5–99.9%
📦 Qubit count (2024): up to 1,121 (IBM)
Trade-offs
✅ Fast gates
✅ Scalable fabrication
✅ Most advanced today
❌ Short coherence vs ion
❌ Needs extreme cooling
❌ Limited connectivity
Why it works: At superconducting temperatures, the Josephson junction behaves like a nonlinear inductor. Its energy levels are quantised — only certain energies are allowed. The two lowest energy levels are |0⟩ and |1⟩. Microwave pulses at the exact transition frequency drive the qubit between states.
⚗️
Wizzy · Quantum Guide
Trapped ion qubits use individual atoms suspended in electromagnetic traps. IonQ and Quantinuum (Honeywell) use ytterbium-171 or barium-133 ions. Qubits are encoded in the ion's internal energy levels. Gates are performed with laser pulses. The remarkable thing: trapped ions maintain coherence for minutes — 10,000× longer than superconducting qubits.
🌀 Individual atoms as computers
Every ion of the same element is perfectly identical — manufactured by nature, not by a fab line. This gives trapped ion qubits naturally high fidelity. The downside: laser operations are slower, and packing many ions together while keeping them individually addressable becomes extremely challenging above ~30–50 qubits per trap.

Trapped Ion Qubits — IonQ & Quantinuum

Key Specs
🌡️ Temperature: ~1 millikelvin
⏱️ Coherence (T2): 1 minute – hours
⚡ Gate time: 1–10 µs (100× slower)
🎯 Gate fidelity: 99.9–99.99%
📦 Qubit count (2024): 32–56 (IonQ Forte)
Trade-offs
✅ Very long coherence
✅ All-to-all connectivity
✅ Highest gate fidelity
❌ Slow gate operations
❌ Hard to scale past ~50
❌ Complex laser systems
All-to-all connectivity: In superconducting processors, qubits are only connected to their nearest neighbours (like squares on a chessboard). In trapped ion systems, any ion can interact with any other ion via shared vibrations of the trap — all-to-all connectivity. This massively reduces the circuit depth needed for many algorithms.
💡
Wizzy · Quantum Guide
Two more approaches with very different bets. Photonic: qubits in photons (particles of light) — naturally room temperature, but photons barely interact with each other, making two-qubit gates extremely challenging. Topological: Microsoft's long-term bet on exotic quantum particles (Majorana anyons) that are theoretically immune to environmental noise.
🌀 Why Microsoft bet on topology
Regular qubits store information locally — one atom, one junction, one photon. Any local disturbance can corrupt it. Topological qubits encode information in a non-local, global property of a quantum system — like a knot in a rope. You can disturb any local part without untying the knot. This topological protection would make qubits inherently more stable. The challenge: creating and controlling Majorana anyons is extraordinarily difficult.

Photonic & Topological Qubits

💡 Photonic (PsiQuantum)
🌡️ Room temperature!
⏱️ Coherence: picoseconds
⚡ Very fast (light speed)
🎯 Single-qubit: excellent
❌ 2-qubit gates: very hard
📦 Strategy: millions of photonic qubits with measurement-based computing
🌀 Topological (Microsoft)
🌡️ Very cold (like SC)
⏱️ Theoretically: very long
🛡️ Intrinsic error protection
❌ Still largely theoretical
❌ First demos: 2023
📦 Strategy: fewer, better qubits — quality over quantity
The broader picture: No technology has "won" the qubit race. Superconducting has the most qubits today. Trapped ion has the best fidelity. Photonic might scale to millions. Topological might need the fewest error correction overheads. Different applications may ultimately prefer different hardware — just as CPUs and GPUs coexist in classical computing.
📊
Wizzy · Quantum Guide
Now compare all four technologies side by side. The radar chart shows each technology's strengths across 5 dimensions. No technology is best at everything — that's why multiple approaches are being pursued simultaneously. Click each technology to highlight it on the chart.

Technology Comparison

TechnologyGate SpeedCoherenceFidelityScaleMaturity
🧲 SuperconductingFastMedHighBest★★★★★
⚗️ Trapped IonSlowBestBestLow★★★★
💡 PhotonicFastestV.ShortMedHigh★★
🌀 TopologicalMedV.LongBest*Med★
* Topological fidelity is theoretical — not yet demonstrated at scale
🧲
Wizzy · Quantum Guide
🎊 You now understand the real physical basis of quantum computing! The hardware challenge is as important as the algorithmic challenge. Great algorithms mean nothing if the qubits decohere before the computation finishes. Session 5 continues with decoherence and error correction — the engineering solutions to keep qubits alive.
🧠 What you actually learned today
  • Superconducting qubits (IBM, Google): Josephson junctions at 15mK. Fast gates, moderate coherence, most qubits today, scalable but limited connectivity.
  • Trapped ion qubits (IonQ, Quantinuum): individual atoms in laser traps. Slow gates but minutes of coherence, highest fidelity, all-to-all connectivity, hard to scale past ~50.
  • Photonic qubits (PsiQuantum): room temperature, light-speed operations, but two-qubit gates are extremely challenging due to photon non-interaction.
  • Topological qubits (Microsoft): Majorana anyons with intrinsic noise protection — still largely theoretical, first demonstrations only in 2023–2024.
  • No single technology "wins" — different trade-offs suit different applications, just as CPUs and GPUs coexist in classical computing.
🧲

Hardware Expert Badge!

You understand what real quantum computers are made of — and why building them is so hard!

🧲 WhizzStep Quantum Lab
This certifies that
Student Name
has mastered Quantum Hardware — Superconducting, Trapped Ion, Photonic & Topological Qubits
Hardware Expert
Qubit Zoo
Real Hardware
📖 Quantum Vocabulary
Josephson junction NEW

A thin insulating barrier between two superconductors. Quantum tunnelling through this barrier creates a nonlinear inductor whose energy levels act as a qubit.

Coherence time KEY

How long a qubit maintains its quantum state before environmental noise destroys it. Superconducting: ~100µs. Trapped ion: minutes. Room temp: nanoseconds.

Like how long a spinning top stays upright before falling over.
Gate fidelity NEW

The probability that a gate operation produces the correct output. 99.9% fidelity means 1 in 1,000 gates makes an error. Need ~99.99% for fault tolerance.

Topological qubit NEW

A qubit encoded in a global topological property of a quantum system — like a knot. Local disturbances cannot corrupt it because they cannot change the global topology.

Millikelvin

One thousandth of a Kelvin. 15mK = 0.015K = -273.135°C. Colder than deep space (2.7K). Required for superconducting quantum processors.

NISQ era

Noisy Intermediate-Scale Quantum — the current era: 50–1000 qubits with significant error rates. Before fault-tolerant quantum computing becomes possible.

Key Concepts from Q13

Hardware race

🏁 No clear winner

IBM has the most qubits. IonQ has the best fidelity. PsiQuantum is betting on silicon photonics. Microsoft is betting on topology. The race is genuinely open.

Trade-offs

⚖️ Speed vs coherence

Fast gates + short coherence (superconducting) vs slow gates + long coherence (trapped ion). The product of gate time and number of gates that fit in the coherence window determines useful circuit depth.

Connectivity

🔗 Not all qubits equal

Superconducting qubits connect only to neighbours. Trapped ions connect to any other ion. Connectivity affects circuit depth — poor connectivity forces extra SWAP gates.

Engineering gap

📏 Physical vs logical

Today: ~1,000 physical qubits. Need: ~4M physical qubits for useful fault-tolerant computation. Error correction overhead is the main challenge — hence Q15.