DIO Tab

The Digital I/O has 3 operation modes: Manual means controlled manually, QA Sequencer n means controlled by QA Sequencer n, QA there are the 32-bit DIO port is in use.

Figure 2 shows the architecture of the DIO port. It features 32 bits that can be configured byte-wise as inputs or outputs by means of a drive signal. The digital output data is latched synchronously with the falling edge of the internal clock, which is running at 50 MHz. The internal sampling clock is available at the DOL pin of the DIO connector. Digital input data can either be sampled by the internal clock or by an external clock provided through the CLKI pin. A decimated version of the input clock is used to sample the input data. The Decimation unit counts the clocks to decimation and then latches the input data. The default decimation is 5625000, corresponding to a digital input sampling rate of 1 sample per second.

In Manual mode, each DIO pin can be controlled manually according to Figure 2 and the DIO interface specification is detailed in Specifications.

In QA Sequencer N mode, each DIO pin output can be controlled in the SHFQA Readout Pulse Generator Tab Sequencer N (N indicates which Channel) by a SeqC command setDIO.

In QA Results QCCS mode, DIO pins are configured to send Readout Result after Thresholding. Table 2 shows the mapping between DIO pins and input/output signals available when using the DIO connector to output qubit state measurement results. The direction is as seen from the SHFQA Instrument. In order to use these signals, the Digital I/O Mode and Drive setting have to be chosen accordingly.

Figure 3 belows shows an example of multiple channel readout transmissions through DIO. Every readout is sent in a single message. A one-hot encoding of the readout channel is sent along with the readout on dedicated bits. The valid bit is set for every valid DIO transaction.

The QA Results QCCS mode of the DIO interface provides one way of communicating discriminated qubit states between 2 Instruments. For more than 2 Instruments, both qubit states and synchronization becomes essential. The following section explains how the ZSync interface works for both Instrument synchronization and feedback. Please note that ZSync settings are under the Device Tab.

The ZSync link of the Zurich Instruments' Quantum Computing Control System (QCCS) enables Instrument synchronization and communication on the system level through the Zurich Instruments' PQSC Programmable Quantum System Controller. This architecture is able to support quantum algorithms run in scalable quantum processors.

In particular, the ZSync links distribute the system clock to all Instruments and synchronize all Instruments to sub-nanosecond levels. Besides status monitoring to ensure quality and reliability of qubit tune-up routines, it provides a bidirectional data interface to send readout results to, or obtain sequence instructions from the PQSC.

The ZSync links adhere to strict real-time behavior: all data transfers are predictable to single clock cycle precision. In the SHFQA, the link is optimized for maximum data transfer bandwidth to the central controller. For example, twice the bandwidth is reserved for results being transfered to the PQSC with respect to the allocated bandwith for instructions that are received from the PQSC. This enables global feedback and error correction through centralized syndrome decoding and synchronized actions on the global QCCS system level.

Using the startQA- command, the SHFQA or the Quantum Analyzer Channel of the SHFQC generates a readout result and forwards it to the PQSC over the ZSync. Depending on the address provided, the PQSC stores it in the register bank - the center of the feedback in the QCCS system. After processing, the PQSC then forwards the results to other devices in the QCCS, such as the SHFSG.

The register bank requires a readout to have an address and a mask along with the readout data. Each component is sent in a separate ZSync message. The address is sent first, followed by the mask, and then the data, see Figure 4. To reduce latency, the address and the mask are sent during the readout, and the data is then sent as soon as the discriminated qubit results are ready.

For the SHFQA with multiple channels, the forwarding of the readout results is interleaved to reduce latency. For example, if all four readout channels perform a readout at the same time, the SHFQA transmits the results of two channels on a single lane of a dual-lane ZSync. On each lane, the SHFQA first sends the header and mask of the two readout channels during the readout. The data is then sent as soon as it is available.