compiler_abc#

Compiler base and utility classes for Qblox backend.

Module Contents#

Classes#

`InstrumentCompiler`	Abstract base class that defines a generic instrument compiler.
`SequencerCompiler`	Class that performs the compilation steps on the sequencer level.
`_ModuleSettingsType`	A typehint for the various module settings (e.g.
`ClusterModuleCompiler`	Base class for all cluster modules, and an interface for those modules to the

Attributes#

`logger`
`_SequencerT_co`	A generic SequencerCompiler type for typehints in `ClusterModuleCompiler`.

logger[source]#

class InstrumentCompiler(name: str, total_play_time: float, instrument_cfg: quantify_scheduler.backends.qblox_backend._ClusterModuleCompilationConfig | quantify_scheduler.backends.qblox_backend._ClusterCompilationConfig | quantify_scheduler.backends.qblox_backend._LocalOscillatorCompilationConfig)[source]#

Bases: abc.ABC

Abstract base class that defines a generic instrument compiler.

The subclasses that inherit from this are meant to implement the compilation steps needed to compile the lists of OpInfo representing the pulse and acquisition information to device-specific instructions.

Each device that needs to be part of the compilation process requires an associated InstrumentCompiler.

Parameters:

name – Name of the QCoDeS instrument this compiler object corresponds to.
total_play_time – Total time execution of the schedule should go on for. This parameter is used to ensure that the different devices, potentially with different clock rates, can work in a synchronized way when performing multiple executions of the schedule.
instrument_cfg – The compilation config referring to this device.

name[source]#

total_play_time[source]#

instrument_cfg[source]#

prepare(**kwargs) → None[source]#: Method that can be overridden to implement logic before the main compilation starts. This step is to extract all settings for the devices that are dependent on settings of other devices. This step happens after instantiation of the compiler object but before the start of the main compilation.

abstract compile(debug_mode: bool, repetitions: int) → object[source]#

An abstract method that should be overridden in a subclass to implement the actual compilation. It should turn the pulses and acquisitions added to the device into device-specific instructions.

Parameters:

debug_mode – Debug mode can modify the compilation process, so that debugging of the compilation process is easier.
repetitions – Number of times execution of the schedule is repeated.

Returns:

A data structure representing the compiled program. The type is dependent on implementation.

class SequencerCompiler(parent: ClusterModuleCompiler, index: int, static_hw_properties: quantify_scheduler.backends.types.qblox.StaticHardwareProperties, sequencer_cfg: quantify_scheduler.backends.qblox_backend._SequencerCompilationConfig)[source]#

Bases: abc.ABC

Class that performs the compilation steps on the sequencer level.

Abstract base class for different sequencer types.

Parameters:

parent – A reference to the module compiler this sequencer belongs to.
index – Index of the sequencer.
static_hw_properties – The static properties of the hardware. This effectively gathers all the differences between the different modules.
sequencer_cfg – The instrument compiler config associated to this instrument.

_settings: quantify_scheduler.backends.types.qblox.SequencerSettings[source]#

parent[source]#

index[source]#

port[source]#

clock[source]#

op_strategies: list[quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy] = [][source]#

_num_acquisitions = 0[source]#

static_hw_properties[source]#

register_manager[source]#

qasm_hook_func[source]#

latency_correction[source]#

distortion_correction[source]#

property connected_output_indices: tuple[int, Ellipsis][source]#

Return the connected output indices associated with the output name specified in the hardware config.

For the baseband modules, output index ‘n’ corresponds to physical module output ‘n+1’.

For RF modules, output indices ‘0’ and ‘1’ (or: ‘2’ and ‘3’) correspond to ‘path_I’ and ‘path_Q’ of some sequencer, and both these paths are routed to the same physical module output ‘1’ (or: ‘2’).

property connected_input_indices: tuple[int, Ellipsis][source]#

Return the connected input indices associated with the input name specified in the hardware config.

For the baseband modules, input index ‘n’ corresponds to physical module input ‘n+1’.

For RF modules, input indices ‘0’ and ‘1’ correspond to ‘path_I’ and ‘path_Q’ of some sequencer, and both paths are connected to physical module input ‘1’.

property portclock: tuple[str, str][source]#

A tuple containing the unique port and clock combination for this sequencer.

Returns:: The portclock.

property settings: quantify_scheduler.backends.types.qblox.SequencerSettings[source]#

Gives the current settings.

Returns:: The settings set to this sequencer.

property name: str[source]#

The name assigned to this specific sequencer.

Returns:: The name.

property has_data: bool[source]#

Whether or not the sequencer has any data (meaning pulses or acquisitions) assigned to it or not.

Returns:: Has data been assigned to this sequencer?

abstract get_operation_strategy(operation_info: quantify_scheduler.backends.types.qblox.OpInfo) → quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy[source]#

Determines and instantiates the correct strategy object.

Parameters:: operation_info – The operation we are building the strategy for.
Returns:: The instantiated strategy object.

add_operation_strategy(op_strategy: quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy) → None[source]#

Adds the operation strategy to the sequencer compiler.

Parameters:: op_strategy – The operation strategy.

_generate_awg_dict() → dict[str, Any][source]#

Generates the dictionary that contains the awg waveforms in the format accepted by the driver.

Notes

The final dictionary to be included in the json that is uploaded to the module is of the form:

program
awg
    waveform_name
        data
        index
acq
    waveform_name
        data
        index

This function generates the awg dictionary.

Returns:

The awg dictionary.

Raises:

ValueError – I or Q amplitude is being set outside of maximum range.
RuntimeError – When the total waveform size specified for a port-clock combination exceeds the waveform sample limit of the hardware.

_generate_weights_dict() → dict[str, Any][source]#

Generates the dictionary that corresponds that contains the acq weights waveforms in the format accepted by the driver.

Notes

The final dictionary to be included in the json that is uploaded to the module is of the form:

program
awg
    waveform_name
        data
        index
acq
    waveform_name
        data
        index

This function generates the acq dictionary.

Returns:: The acq dictionary.
Raises:: NotImplementedError – Currently, only two one dimensional waveforms can be used as acquisition weights. This exception is raised when either or both waveforms contain both a real and imaginary part.

_validate_awg_dict(wf_dict: dict[str, Any]) → None[source]#

abstract _prepare_acq_settings(acquisitions: list[quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy], acq_metadata: quantify_scheduler.schedules.schedule.AcquisitionMetadata) → None[source]#

Sets sequencer settings that are specific to certain acquisitions. For example for a TTL acquisition strategy.

Parameters:

acquisitions – List of the acquisitions assigned to this sequencer.
acq_metadata – Acquisition metadata.

_generate_acq_declaration_dict(repetitions: int, acq_metadata: quantify_scheduler.schedules.schedule.AcquisitionMetadata) → dict[str, Any][source]#

Generates the “acquisitions” entry of the program json. It contains declaration of the acquisitions along with the number of bins and the corresponding index.

For the name of the acquisition (in the hardware), the acquisition channel (cast to str) is used, and is thus identical to the index. Number of bins is taken to be the highest acq_index specified for that channel.

Parameters:

repetitions – The number of times to repeat execution of the schedule.
acq_metadata – Acquisition metadata.

Returns:

The “acquisitions” entry of the program json as a dict. The keys correspond to the names of the acquisitions (i.e. the acq_channel in the scheduler).

generate_qasm_program(ordered_op_strategies: list[quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy], total_sequence_time: float, align_qasm_fields: bool, acq_metadata: quantify_scheduler.schedules.schedule.AcquisitionMetadata | None, repetitions: int) → str[source]#

Generates a QASM program for a sequencer. Requires the awg and acq dicts to already have been generated.

Example of a program generated by this function:

        wait_sync     4
        set_mrk       1
        move          10,R0         # iterator for loop with label start
start:
        wait          4
        set_awg_gain  22663,10206  # setting gain for 9056793381316377208
        play          0,1,4
        wait          176
        loop          R0,@start
        set_mrk       0
        upd_param     4
        stop

Parameters:

ordered_op_strategies – A sorted list of operations, in order of execution.
total_sequence_time – Total time the program needs to play for. If the sequencer would be done before this time, a wait is added at the end to ensure synchronization.
align_qasm_fields – If True, make QASM program more human-readable by aligning its fields.
acq_metadata – Acquisition metadata.
repetitions – Number of times to repeat execution of the schedule.

Returns:

The generated QASM program.

Warns:

RuntimeWarning – When number of instructions in the generated QASM program exceeds the maximum supported number of instructions for sequencers in the type of module.

Raises:

RuntimeError – Upon total_sequence_time exceeding QASMProgram.elapsed_time.

_assert_total_play_time_on_nco_grid() → None[source]#

Raises an error if the total play time does not align with the NCO grid time.

Method is implemented on the base class instead of the AnalogSequencerCompiler subclass because it is called by generate_qasm_program.

class ParseOperationStatus[source]#

Bases: enum.Enum

Return status of the stack.

COMPLETED_ITERATION[source]#: The iterator containing operations is exhausted.

EXITED_CONTROL_FLOW[source]#: The end of a control flow scope is reached.

_parse_operations(operations_iter: Iterator[quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy], qasm: quantify_scheduler.backends.qblox.qasm_program.QASMProgram, acquisition_multiplier: int) → ParseOperationStatus[source]#: Handle control flow and insert Q1ASM.

abstract _insert_qasm(op_strategy: quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy, qasm_program: quantify_scheduler.backends.qblox.qasm_program.QASMProgram) → None[source]#: Get Q1ASM instruction(s) from op_strategy and insert them into qasm_program.

abstract _write_pre_wait_sync_instructions(qasm: quantify_scheduler.backends.qblox.qasm_program.QASMProgram) → None[source]#

Write instructions to the QASM program that must come before the first wait_sync.

The duration must be equal for all module types.

abstract _write_repetition_loop_header(qasm: quantify_scheduler.backends.qblox.qasm_program.QASMProgram) → None[source]#

Write the Q1ASM that should appear at the start of the repetition loop.

The duration must be equal for all module types.

_get_ordered_operations() → list[quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy][source]#: Get the class’ operation strategies in order of scheduled execution.

_initialize_append_mode_registers(qasm: quantify_scheduler.backends.qblox.qasm_program.QASMProgram, op_strategies: list[quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy]) → None[source]#

Adds the instructions to initialize the registers needed to use the append bin mode to the program. This should be added in the header.

Parameters:

qasm – The program to add the instructions to.
op_strategies – An operations list including all the acquisitions to consider.

_get_latency_correction_ns(latency_correction: float) → int[source]#

_remove_redundant_update_parameters() → None[source]#

Removing redundant update parameter instructions. If multiple update parameter instructions happen at the same time, directly after each other in order, then it’s safe to only keep one of them.

Also, real time io operations act as update parameter instructions too. If a real time io operation happen ((just after or just before) and at the same time) as an update parameter instruction, then the update parameter instruction is redundant.

_validate_update_parameters_alignment() → None[source]#

static _replace_digital_pulses(op_strategies: list[quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy]) → list[quantify_scheduler.backends.qblox.operation_handling.base.IOperationStrategy][source]#: Replaces MarkerPulse operations by explicit high and low operations.

static _generate_waveforms_and_program_dict(program: str, waveforms_dict: dict[str, Any], weights_dict: dict[str, Any] | None = None, acq_decl_dict: dict[str, Any] | None = None) → dict[str, Any][source]#

Generates the full waveforms and program dict that is to be uploaded to the sequencer from the program string and the awg and acq dicts, by combining them and assigning the appropriate keys.

Parameters:

program – The compiled QASM program as a string.
waveforms_dict – The dictionary containing all the awg data and indices. This is expected to be of the form generated by the generate_awg_dict method.
weights_dict – The dictionary containing all the acq data and indices. This is expected to be of the form generated by the generate_acq_dict method.
acq_decl_dict – The dictionary containing all the acq declarations. This is expected to be of the form generated by the generate_acq_decl_dict method.

Returns:

The combined program.

static _dump_waveforms_and_program_json(wf_and_pr_dict: dict[str, Any], label: str | None = None) → str[source]#

Takes a combined waveforms and program dict and dumps it as a json file.

Parameters:

wf_and_pr_dict – The dict to dump as a json file.
label – A label that is appended to the filename.

Returns:

The full absolute path where the json file is stored.

prepare() → None[source]#: Perform necessary operations on this sequencer’s data before compile() is called.

compile(sequence_to_file: bool, align_qasm_fields: bool, repetitions: int = 1) → tuple[dict[str, Any] | None, quantify_scheduler.schedules.schedule.AcquisitionMetadata | None][source]#

Performs the full sequencer level compilation based on the assigned data and settings. If no data is assigned to this sequencer, the compilation is skipped and None is returned instead.

Parameters:

sequence_to_file – Dump waveforms and program dict to JSON file, filename stored in SequencerCompiler.settings.seq_fn.
align_qasm_fields – If True, make QASM program more human-readable by aligning its fields.
repetitions – Number of times execution the schedule is repeated.

Returns:

The compiled program and the acquisition metadata. If no data is assigned to this sequencer, the compilation is skipped and None is returned instead.

_SequencerT_co[source]#

A generic SequencerCompiler type for typehints in ClusterModuleCompiler.

Covariant so that subclasses of ClusterModuleCompiler can use subclassses of SequencerCompiler in their typehints.

class _ModuleSettingsType[source]#

Bases: Protocol

A typehint for the various module settings (e.g. BasebandModuleSettings) classes.

to_dict() → dict[str, Any][source]#: Convert the settings to a dictionary.

class ClusterModuleCompiler(name: str, total_play_time: float, instrument_cfg: quantify_scheduler.backends.qblox_backend._ClusterModuleCompilationConfig)[source]#

Bases: InstrumentCompiler, Generic[_SequencerT_co], abc.ABC

Base class for all cluster modules, and an interface for those modules to the ClusterCompiler.

This class is defined as an abstract base class since the distinctions between the different devices are defined in subclasses. Effectively, this base class contains the functionality shared by all Qblox devices and serves to avoid repeated code between them.

Parameters:

name – Name of the QCoDeS instrument this compiler object corresponds to.
total_play_time – Total time execution of the schedule should go on for. This parameter is used to ensure that the different devices, potentially with different clock rates, can work in a synchronized way when performing multiple executions of the schedule.
instrument_cfg – The instrument compiler config referring to this device.

_settings: _ModuleSettingsType[source]#

instrument_cfg: quantify_scheduler.backends.qblox_backend._ClusterModuleCompilationConfig[source]#

_op_infos: dict[tuple[str, str], list[quantify_scheduler.backends.types.qblox.OpInfo]][source]#

portclock_to_path[source]#

sequencers: dict[str, _SequencerT_co][source]#

property portclocks: list[str][source]#: Returns all the port-clock combinations that this device can target.

property supports_acquisition: bool[source]#

Abstractmethod:

Specifies whether the device can perform acquisitions.

property max_number_of_instructions: int[source]#

Abstractmethod:

The maximum number of Q1ASM instructions supported by this module type.

add_op_info(port: str, clock: str, op_info: quantify_scheduler.backends.types.qblox.OpInfo) → None[source]#

Assigns a certain pulse or acquisition to this device.

Parameters:

port – The port this waveform is sent to (or acquired from).
clock – The clock for modulation of the pulse or acquisition. Can be a BasebandClock.
op_info – Data structure containing all the information regarding this specific pulse or acquisition operation.

property _portclocks_with_data: set[tuple[str, str]][source]#

All the port-clock combinations associated with at least one pulse and/or acquisition.

Returns:: A set containing all the port-clock combinations that are used by this InstrumentCompiler.

property static_hw_properties: quantify_scheduler.backends.types.qblox.StaticHardwareProperties[source]#

Abstractmethod:

The static properties of the hardware. This effectively gathers all the differences between the different modules.

_construct_all_sequencer_compilers() → None[source]#

Constructs SequencerCompiler objects for each port and clock combination belonging to this device.

Raises:: ValueError – Attempting to use more sequencers than available.

abstract _construct_sequencer_compiler(index: int, sequencer_cfg: quantify_scheduler.backends.qblox_backend._SequencerCompilationConfig) → _SequencerT_co[source]#: Create the sequencer object of the correct sequencer type belonging to the module.

distribute_data() → None[source]#: Distributes the pulses and acquisitions assigned to this module over the different sequencers based on their portclocks. Raises an exception in case the device does not support acquisitions.

compile(debug_mode: bool, repetitions: int = 1, sequence_to_file: bool | None = None) → dict[str, Any][source]#

Performs the actual compilation steps for this module, by calling the sequencer level compilation functions and combining them into a single dictionary.

Parameters:

debug_mode – Debug mode can modify the compilation process, so that debugging of the compilation process is easier.
repetitions – Number of times execution the schedule is repeated.
sequence_to_file – Dump waveforms and program dict to JSON file, filename stored in SequencerCompiler.settings.seq_fn.

Returns:

The compiled program corresponding to this module. It contains an entry for every sequencer under the key “sequencers”, and acquisition metadata under the key “acq_metadata”, and the “repetitions” is an integer with the number of times the defined schedule is repeated. All the other generic settings are under the key “settings”. If the device is not actually used, and an empty program is compiled, None is returned instead.