Quantum in the Cloud: Accessing the Freezer

Quantum in the Cloud: Accessing the Freezer

Why you don't need a lab. Learn how to run code on IBM, Amazon, and Google's quantum machines from your bedroom.

The Mainframe Model Returns

In the 1960s, people didn't own computers; they sent "Job Requests" to a central mainframe. Quantum computing is exactly the same today.

Because quantum computers are huge, frozen, and need a team of PhDs to maintain, you don't buy one. You rent one.


1. The Major Portals

Three companies currently dominate the space of "Quantum as a Service":

  1. IBM Quantum: The pioneer. They provide a web interface where you can drag and drop gates onto a circuit or send Python code. Many of their smaller machines are Free for the public.
  2. AWS Braket (Amazon): A "Super-Portal." Instead of just Google or IBM, Braket lets you access hardware from different companies like IonQ, Rigetti, and Oxford Quantum Circuits using one common language.
  3. Azure Quantum (Microsoft): Similar to AWS, focusing on "Resource Estimation"—helping you calculate how many qubits you would need before you spend the money.

2. How the Request Works

When you "Run" a quantum program, you aren't actually connected "Live" to the Qubit.

  1. You send your Circuit (The instructions) to the cloud.
  2. The cloud provider puts your request in a Queue.
  3. When it's your turn, the machine "runs" your circuit a few hundred times (called Shots) to get a statistical average.
  4. The cloud sends the Results (the 0s and 1s) back to your computer.

3. The Cost of Quantum

Quantum computing is currently expensive for enterprises:

  • Reservation Fee: Often ~$1.00 USD just to start the run.
  • Per-Shot Fee: A tiny fraction of a cent per execution.
  • Total: A complex experiment can cost anywhere from $10 to $10,000 depending on the priority and machine quality.

Tip: Always use a simulator first!

sequenceDiagram
    participant User as Your Laptop
    participant Cloud as AWS/IBM Cloud
    participant QPU as The Quantum Machine
    User->>Cloud: Send Python Circuit
    Cloud->>Cloud: Queue Management
    Cloud->>QPU: Fire Pulse Sequences
    QPU-->>Cloud: Measure Results (Shots)
    Cloud-->>User: Return Probability Graph

4. Summary: The Democratization of Physics

This is the most exciting part of the field: The barriers to entry are gone. A college student in Mumbai has the same access to a cutting-edge IBM quantum processor as a researcher at MIT. All you need is an API key.


Exercise: The "Online Order" Analogy

  1. Running a quantum circuit is like ordering Custom Furniture.
  2. You don't go to the factory.
  3. You send the Blueprint (Circuit) to the factory.
  4. They put your blueprint in the Production Line (Queue).
  5. They build the table (Execution) and send you a Photo of it (Results).
  6. You don't need to know how to use a saw; you just need to know how to draw the blueprint!

What's Next?

Quantum computers aren't working alone. They are actually part of a team. In the next lesson, we’ll explore the Hybrid Cloud—the role of classical computers in a quantum system.

Subscribe to our newsletter

Get the latest posts delivered right to your inbox.

Subscribe on LinkedIn