Hacker News new | ask | show | jobs
by greeneggs 3112 days ago
They have both a simulator and real machines you can use. The "ibmqx5" is a 16-qubit device. The problem is that noise rates are very high. For example, just to swap two adjacent qubits (using three CNOT gates), it looks like you will incur at least 10% error. It is hard (though not impossible) to run meaningful experiments on such a noisy system.

https://quantumexperience.ng.bluemix.net/qx/editor

1 comments

Oh, thanks. You are correct that it is very noisy. Just a single CNOT is giving me 15% error.
Here are the numbers for their latest 20-qubit device. CNOT error rates are a bit lower. I don't know if they plan on making this publicly accessible.

https://youtu.be/T-8uuq7Izl8?t=26m36s