Closed singular-value closed 3 years ago
There was internal error at Rigetti that caused the failure. We have alerted them and they are investigating. I will update this as we get more information.
The Aspen-9 is running tasks again. If you would like to resubmit your tasks, we can see if the error is recreated or if the issue recurs.
Rigetti has fixed the bug on their end. Please confirm if you are able to run your ansatz.
I'm also wondering if we were still charged for the failed runs, or if they don't count against us since they didn't actually execute.
No. You will only be charged for COMPLETED tasks, not FAILED or CANCELLED tasks. There are additional charges for simulator and notebook instance runtime as well. https://aws.amazon.com/braket/pricing/
Thanks for looking into this @jheckey.
We don't need to test the ansatz anymore—will go ahead and close this issue.
Describe the bug I submitted six circuits to IonQ and to Rigetti. All six circuits are very similar-->same ansatz, but different variational parameters.
Failed to run task; Timeout on lattice Aspen-9; Unable to complete Rigetti compilation; the circuit may be too large
.Failed to run task; lattice: Aspen-9
To reproduce Circuit 1 and 2 are printed below. Note that they are extremely similar-->just different rotation angles:
Circuit 2:
Here the respective repr's for reconstructing the circuits:
and
Expected behavior Given that circuit 1 ran on Rigetti, it's weird that circuits 2-6 failed to run. In addition, all 6 circuits ran on IonQ. Perhaps there is some compilation timeout that triggered on circuit 2 and then carried over to circuit 6? (but only for Rigetti?)
Screenshots or logs Task IDs for the circuits for Rigetti, IonQ:
System information A description of your system. Please provide:
Additional context I'm also wondering if we were still charged for the failed runs, or if they don't count against us since they didn't actually execute.