6
$\begingroup$

IBM Q is (maybe first) quantum computing platform allowing users to register for free and run algorithm on real quantum processors. There are several processors with usually 5 qubits, however, it is also possible to use processor Melbourne with 15 qubits.

My question: Is there any other platform where it is possible to register for free and run programs on real quantum hardware?

$\endgroup$
  • $\begingroup$ all good, I'm curious what will come. I know ionq.com but I'm not sure if it free. I'll remove my comments... $\endgroup$ – draks ... Apr 29 at 10:49
5
$\begingroup$

Already mentioned is:

  • IBM Quantum Experience, the widest known and largest platform for small quantum computing experiments. There are various physical backends containing upwards of 20 qubits for free users, and more for 'IBM Q partners'. The chips have varying qubit connectivity (architecture), which can be of vital importance to low-level experiments. Although there is a browser-based interface for easy access, the main interface is the IBM Q Experience through qiskit. As qiskit is developed by IBM researchers, this offers the go-to way to interface with the hardware.
  • Quantum Inspire by QuTech, a joint venture by Delft University of Technology and TNO. Quantum inspire offers both a web-based interface based on the cQASM language, an interface through Qiskit and an interface through ProjectQ, both including a transpiler and compiler. There are currently 2 physical chips, one containing two and one containing 5 qubits.

Both of these platforms require a free account and use queue-based token-restricted access system. More tokens can often be requested by researchers and hobbyist with good reason.

I'm also familiar with:

  • Rigetti's Forest platform, which works on a request-only based access system. As far as I know, access is always granted to researchers with a good proposal. The physical backends offered contain up to 28 qubits. Interfacing is done using the Forest SDK containing pyQuil.
  • A different approach: D-Wave's software platform offers (I believe) access to their adiabatic QPU's. I have never worked with this and am far from an expert on the annealing method of quantum computing. Note that this is an entirely different approach to quantum computing than the usual gate-based model used prominently on this website. See this excellent set of answers on this website by Neil de Beaudrap for an overview of all the different models.

I am sure there are more platforms that I am not familiar with!

| improve this answer | |
$\endgroup$
  • $\begingroup$ Thanks for mentioning Rigetti and D-Wave. $\endgroup$ – Martin Vesely Apr 29 at 10:43
6
$\begingroup$

There is a platform Quantum Inspire provided by QuTech, an organization co-founded by Delft University of Technology. Currently, the platform offers two real quantum processors:

  • Spin-2 - 2 qubits processor with qubits implemented as electron spins
  • Starmon-5 - 5 qubits transmon processor

The platform uses a programming language based on QASM. It is also possible to use Qiskit.

Here is a link to the platform.

In comparison with IBM Q, decoherence is more or less similar for transmon processors, while spin processor of Quantum Inspire shows higher decoherence. I tried to implement a circuit for preparing Bell state $\frac{1}{\sqrt{2}}(|01\rangle + |10\rangle)$. On IBM Q (processor Rome), I reached these results (4,096 shots on both platforms):

  • $|00\rangle$ - 4.13 %
  • $|01\rangle$ - 50.68 %
  • $|10\rangle$ - 43.29 %
  • $|11\rangle$ - 1.9 %

On Quantum Inspire, the results were these

  • $|00\rangle$ - Starmon-5: 9.1 %, Spin-2: 9.1 %
  • $|01\rangle$ - Starmon-5: 44.2 %, Spin-2: 16.8 %
  • $|10\rangle$ - Starmon-5: 43.4 %, Spin-2: 60.5 %
  • $|11\rangle$ - Starmon-5: 3.2 %, Spin-2: 13.6 %
| improve this answer | |
$\endgroup$
  • 1
    $\begingroup$ I highly recommend to use qiskit + the quantum inspire SDK (pip install quantuminsipre; github.com/QuTech-Delft/quantuminspire) over the browser based cQASM interface for anything more than very small proof-of-concept experiments! $\endgroup$ – JSdJ Apr 29 at 9:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.