THE QUANTUM COMPUTING LANDSCAPE TODAY
Who is building what — IBM, Google, IonQ, D-Wave, and how you can access real quantum hardware right now.
Quantum computing isn't a single company's project or a single technology — it's an entire ecosystem of competing hardware approaches, cloud platforms, and companies making different bets on how to get to a useful, large-scale quantum computer. This lesson gives you a map of that ecosystem as it stands today, so that names like "IonQ" or "superconducting qubit" have a home in your head before Module 3 dives into the hardware details.
The Major Players and Their Bets
Different companies are betting on fundamentally different physical technologies to build a qubit. You'll get the full engineering detail on each of these in Module 3 — this lesson is just the map, not the territory.
| COMPANY | QUBIT TECHNOLOGY | NOTABLE FACT |
|---|---|---|
| IBM | Superconducting circuits | Public roadmap targeting 100,000+ qubit systems by 2033; runs the IBM Quantum cloud platform used throughout Module 11 |
| Superconducting circuits | Made the 2019 "quantum supremacy" claim with its 53-qubit Sycamore processor | |
| IonQ | Trapped ions | Publicly traded; known for high gate fidelity (accuracy) compared to superconducting approaches |
| Quantinuum | Trapped ions | Formed from the merger of Honeywell Quantum Solutions and Cambridge Quantum |
| Rigetti | Superconducting circuits | Publicly traded; focuses on cloud-accessible quantum computing |
| PsiQuantum | Photonic (light-based) qubits | Aiming to build a fault-tolerant, million-qubit machine using silicon photonics |
| D-Wave | Quantum annealing (not gate-model) | Sold the first commercial quantum computer in 2011; covered separately in Module 22 |
| Microsoft | Topological qubits (experimental) | Betting on a qubit type that is inherently more resistant to noise, still largely unproven at scale |
Nobody has definitively "won" the hardware race yet — each technology has different tradeoffs between qubit quality, how many qubits you can fit together, and how hard the engineering is to scale up. Module 3 covers this comparison (coherence, fidelity, scalability) in full detail. For now, just know that the field is still in an experimental, multi-approach phase, similar to how early classical computing experimented with vacuum tubes, relays, and mechanical switches before settling on transistors.
Cloud Access: You Don't Need to Own a Quantum Computer
Here's the practically important fact for you as a learner: every major quantum hardware provider offers cloud access, meaning you can write a quantum program on your laptop and run it on a real quantum computer over the internet, often for free at a small scale.
- IBM Quantum — free tier access to real superconducting hardware, directly usable from the Qiskit framework you'll learn in Module 11.
- Amazon Braket — a cloud aggregator giving access to multiple hardware providers (including IonQ and Rigetti) through one unified interface.
- Microsoft Azure Quantum — another aggregator, bundling access to several hardware partners.
- IonQ and Quantinuum — also offer direct cloud access to their own trapped-ion systems.
This is exactly the model you'll use in Module 11, Lesson 7 ("Running on Real IBM Quantum Hardware"), where you'll take a circuit you built and simulated locally, and send it to run on an actual physical quantum chip.
What "Qubit Count" Does and Doesn't Tell You
You'll frequently see headlines like "Company X announces a 1,000-qubit processor." It's tempting to treat qubit count like clock speed — bigger number, better computer — but this is misleading for two reasons you'll fully understand once you reach Part C of this curriculum:
- Not all qubits are equally reliable. A 1,000-qubit machine with high error rates can be less useful than a 50-qubit machine with very low error rates, because errors compound as circuits get longer and more qubits are involved (Module 16).
- Connectivity matters. Some hardware only allows neighboring qubits to interact directly, which affects what circuits can run efficiently (Module 18 covers this in depth, under "qubit connectivity and coupling maps").
Fidelity, loosely, measures how close a real physical qubit's actual behavior is to its ideal, error-free behavior — expressed as a percentage (e.g., 99.9% gate fidelity). Higher fidelity means fewer errors per operation. This is a term you'll see constantly once you reach Module 3's hardware comparison and Module 16's noise discussion — for now, just know that "more qubits" and "better qubits" are two separate, both-important axes, not the same thing.
When you read a quantum computing headline, ask three questions: How many qubits? What's the error rate/fidelity? And is this solving a real problem, or a deliberately chosen benchmark task? All three matter — qubit count alone tells you very little.
Practice Questions
Test your understanding
KEY TAKEAWAYS
Remember these points
Multiple companies are pursuing different physical qubit technologies (superconducting, trapped ion, photonic, topological, annealing) with no single clear winner yet
Cloud platforms (IBM Quantum, Amazon Braket, Azure Quantum) let you run real quantum programs on physical hardware from your laptop, often for free at small scale — this is the exact model Module 11 uses
Qubit count alone is a misleading headline metric — fidelity (error rate) and connectivity matter just as much for real-world usefulness
A healthy skepticism toward "bigger qubit count" headlines will serve you well as you follow the field's progress