You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As a more systematic solution to the problem exposed in #1062, another option is to allow extensions of the Qibolab native instructions, such that they can be developed separately from the main package.
This would make it possible to essentially factor pieces of low-level code in the same way it happens for functions in programming languages.
The feature would require both a mechanism for listing new instructions (pulse-like objects, for PulseSequences), and a way to extend the driver to process the given instruction, turning it into the required device's input (i.e. machine assembly code and configs, or calls to the provided API, according to the driver internal structure).
There should be a standard mechanism for drivers to retrieve the possibly registered implementation for non-native instructions.
Example
For the case of the on device RB (which may implemented on top of different lower level primitives, like sampling), there could be an RB(depth) instruction/pulse-like, and this may be unraveled into the code required to execute the RB and acquire its results.
The text was updated successfully, but these errors were encountered:
As a more systematic solution to the problem exposed in #1062, another option is to allow extensions of the Qibolab native instructions, such that they can be developed separately from the main package.
This would make it possible to essentially factor pieces of low-level code in the same way it happens for functions in programming languages.
The feature would require both a mechanism for listing new instructions (pulse-like objects, for
PulseSequence
s), and a way to extend the driver to process the given instruction, turning it into the required device's input (i.e. machine assembly code and configs, or calls to the provided API, according to the driver internal structure).There should be a standard mechanism for drivers to retrieve the possibly registered implementation for non-native instructions.
Example
For the case of the on device RB (which may implemented on top of different lower level primitives, like sampling), there could be an
RB(depth)
instruction/pulse-like, and this may be unraveled into the code required to execute the RB and acquire its results.The text was updated successfully, but these errors were encountered: