David/iqm5q march 2024 #126
Conversation
for more information, see https://pre-commit.ci
andrea-pasquale
added
run-on-iqm5q
iqm5q/platform.py
Outdated
| @@ -152,10 +152,11 @@ def create(): | |||
|
|
|||
| # flux | |||
| for i in range(6, 11): | |||
| channels[f"L4-{i}"].power_range = 0.8 | |||
| channels[f"L4-{i}"].power_range = 9 | |||
There was a problem hiding this comment.
This is not a valid value for power range for this instrument. This is HDAWG, and for it the power range seems to be in volts (while for the other instruments it is DB). The valid values are 200mV, 400mV, 600mV, 800mV, 1V, 2V, 3V, 4V, 5V.
For testing my changes in qiboteam/qibolab#848 I used 1V, and things worked perfectly fine.
iqm5q/platform.py
Outdated
| # flux couplers | ||
| for i in range(11, 15): | ||
| channels[f"L4-{i}"].power_range = 0.8 | ||
| channels[f"L4-{i}"].power_range = 9 |
|
@hay-k thanks for the review. I set 9 because I could not find in the documentation what was the maximum value accepted in the PR and it is rounded to the higher allowed value. So, thanks for the comment, I will write a comment in this platform section for future characterization as I did with the PR values for the drive for example. I will set the flux PR to 5, while trying to characterize the coupler because we need to shift the high freq qubit to put it near the low qubit freq. As you know, right now the qubits are not shifting with the flux amplitude so, until them, I think we need to amplify as much as possible the bias to observe a shift. |
DavidSarlle
requested review from
hay-k and
GabrielePalazzo
and removed request for
stavros11,
igres26 and
aorgazf
|
I have checked again this runcard after the latest warm up and cool down and seems to be good. All the fidelities for single qubit are more or less the same (in fact are better for q1 95% and q2 85%). So, from my side this runcard can be merged into main. |
|
Refined all single qubit characterization: Single shot classify (all qubits together) decreases a little bit as expected: If the classifier is run qubit by qubit the fidelities are much better as expected: q0 83% From my side this runcard is ready to be merged into main. |
for more information, see https://pre-commit.ci
Latest chracterization of iqm5q chip (1q gates).
When all the qubits are operated at the same time the fidelities decrease a little bit until we chracterized the couplers: