ARTIQ (Advanced Real-Time Infrastructure for Quantum physics) is a next-generation control system for quantum information experiments. It is being developed in partnership with the Ion Storage Group at NIST, and its applicability reaches beyond ion trapping.

Modern research on quantum information systems poses particular challenges to the control system:

• dozens of signals need to be received and generated with extremely precise timing, in particular to ensure phase coherence
• quantum error correction schemes require very low reaction latency in response to a measurement
• real-world implementations of quantum gates, and a fortiori quantum algorithms, involve structurally complex protocols
• ever-improving experimental techniques drive the need for a flexible and programmable system
• the diversity of equipment, device drivers and data analysis software involved in a single experiment results in a distributed and multi-platform environment

Enter ARTIQ

ARTIQ features a high-level programming language, based on Python, that helps describing complex experiments. It is compiled and executed on dedicated FPGA hardware with nanosecond timing resolution and sub-microsecond latency.

The time-critical code (a kernel) running on the FPGA (the core device) is easily interfaced with Python code on the computer using a remote procedure call (RPC) mechanism.

The FPGA design is highly portable so that it can adapt to different laboratory setups and resist hardware obsolescence.

ARTIQ drivers for non-realtime devices can be run on remote machines with different operating systems.

The project also includes a graphical user interface, an experiment scheduling system, and databases for experiments, devices, parameters and results.

Technologies employed include Python, Migen, MiSoC/mor1kx, LLVM and llvmlite.

A taste of the programming language

Timing language basics

trigger.sync()               # wait for trigger input
start = now()                # capture trigger time
for i in range(3):
    delay(5*us)
    dds.pulse(900*MHz, 7*us) # first pulse 5 µs after trigger
at(start + 1*ms)             # re-reference time-line
dds.pulse(200*MHz, 11*us)    # exactly 1 ms after trigger

• now(), at(), delay() describe time-line of an experiment
• Exact time is kept in an internal variable
• That variable only loosely tracks the execution time of CPU instructions
• The value of that variable is exchanged with the real-time input/output (RTIO) core that does precise timing

Parallelism

with sequential:
    with parallel:
        a.pulse(100*MHz, 10*us)
        b.pulse(200*MHz, 20*us)
    with parallel:
        c.pulse(300*MHz, 30*us)
        d.pulse(400*MHz, 20*us)

• Experiments are inherently parallel: simultaneous laser pulses, parallel cooling of ions in different trap zones
• parallel and sequential contexts with arbitrary nesting
• a and b pulses both start at the same time
• c and d pulses both start when a and b are both done (after 20 µs)
• Currently implemented by inlining, loop-unrolling, and interleaving (i.e. static scheduling at compile time)

Organize experiment components and reuse code

class Experiment(AutoDB):
    def build(self):
        self.ion1 = Ion(...)
        self.ion2 = Ion(...)
        self.transporter = Transporter(...)

    @kernel
    def run(self):
        with parallel:
            self.ion1.cool(duration=10*us)
            self.ion2.cool(frequency=...)
        self.transporter.move(speed=...)
        delay(100*ms)
        self.ion1.detect(duration=...)

Is ARTIQ right for my lab?

If you don't mind getting your hands dirty (i.e. compiling software and FPGA bitstreams), it's easy to find out by yourself! The core device is based on the resource-efficient MiSoC system-on-chip design, and you can use a limited version of ARTIQ with the small and low-cost Papilio Pro board. Feel free to send feedback to the mailing list!

Note that ARTIQ is still in development and we expect to deploy a fully functional release (that will include a GUI) at NIST by June 2015. We are also working on making the installation procedure easier.

Resources

• Source code repository
• Public mailing list
• Slide-show
• Manual (work in progress)
• Commercial support available - contact sb at m-labs.hk