Open igres26 opened 5 days ago
Thanks for reporting this.
I don’t think we will address it as we avoid implementing new features for qibolab 0.1.
In 0.2 there is no start
(and relative_start
) in pulses anymore, as it has been replaced by the Delay
object. Also, all native gates are defined as PulseSequence
s which can also contain Delay
s. Therefore, I would expect that the issue is already addressed there, even though it is not explicitly tested.
I don’t think we will address it as we avoid implementing new features for qibolab 0.1.
Exactly.
Therefore, I would expect that the issue is already addressed there, even though it is not explicitly tested.
We're generally lacking tests on hardware, especially for specific cases. But yes, once the Delay
is there, it is just a matter of honoring it in the drivers' implementation, which could be already tested on QM (OPX+ and OPX1000). From the point of view of the general Qibolab support, the use case is just supported, since any Delay
in a sequence will be taken into account as a valid delay.
Could we rely on you for testing it with QM? @igres26
All the gates have a "relative start" which I feel is barely used.
However, when we calibrate Two-Qubit gates, we are allowed to play around with a delay after the gate is applied. It helps with the virtual phases for example. But there is no way to add a delay after a gate is applied into the parameters of the gate, making these routines that delay on this delay
dt
a bit useless.Could we add on the create CZ module an input for a delay after the pulse is applied?