In Sep 2020, IBM announced its roadmap to reaching 1,000-plus qubits by 2023 to reach the true quantum industry inflection point of Quantum Advantage — the point at which quantum systems will be more powerful than today’s conventional computing. It’s an aggressively ambitious mission, which recognizes that merely incremental increases in qubit counts and more sophisticated algorithms alone will not deliver Quantum Advantage—the point where certain information processing tasks can be performed more efficiently or cost effectively on a quantum computer, versus a classical one.
The roadmap, announced at the annual IBM Quantum Summit, aims to take the technology from today’s noisy, small-scale devices to the million-plus qubit devices of the future. (A qubit is the basic unit of quantum information, analogous to the bits of classical computing.) Such progress is essential if quantum computers are to help industry and research organizations tackle some of the world’s biggest challenges, across industry, government and research.
The IBM Quantum team builds quantum processors—computer processors that rely on the mathematics of elementary particles in order to expand our computational capabilities, running quantum circuits rather than the logic circuits of digital computers. “We represent data using the electronic quantum states of artificial atoms known as superconducting transmon qubits, which are connected and manipulated by sequences of microwave pulses in order to run these circuits. But qubits quickly forget their quantum states due to interaction with the outside world. The biggest challenge facing our team today is figuring out how to control large systems of these qubits for long enough, and with few enough errors, to run the complex quantum circuits required by future quantum applications.”
One of the great challenges for scientists seeking to harness the power of quantum computing is controlling or removing quantum decoherence – the creation of errors in calculations caused by interference from factors such as heat, electromagnetic radiation, and material defects. The errors are especially acute in quantum machines, since quantum information is so fragile. Quantum processors require special conditions to operate, and they must be kept at near-absolute zero, like IBM’s quantum chips are kept at 15mK.
The deep complexity and the need for specialised cryogenics is why at least IBM’s quantum computers are accessible via the cloud, and will be for the foreseeable future, Dasgupta, who is also IBM’s CTO for South Asia region, noted. To maximise the potential of quantum computers, the industry must solve challenges from the cryogenics, production and effects materials at very low temperatures. This is one of the reasons why IBM built its super-fridge to house Condor, IBM Research’s Director Gargi Dasgupta explained. As a solution to solve all these problems at once, cryogen- free dilution refrigerators with mechanical refrigerators in place of liquid helium have been developed.
IBM is already preparing a jumbo liquid-helium refrigerator, or cryostat, to hold a quantum computer with 1 million qubits. Goldeneye is IBM’s internal codename for the world’s largest dilution refrigerator, which will house a future 1,000,000 qubit quantum processor. The IBM road map doesn’t specify when such a machine could be built. But if company researchers really can build a 1000-qubit computer in the next 2 years, that ultimate goal will sound far less fantastical than it does now.
Superconducting qubits need to be cooled down to between 10-15 millikelvin for their quantum behavior to emerge. They need to be kept that cold to ensure that their performance is high. Dilution refrigeration technology, which has been around for a really long time, is an enabling technology specifically for superconducting qubits for quantum computing. Whereas a different type of qubit might require its own unique set of hardware and infrastructure.
Around 2010, cryogen-free dilution refrigerators became en vogue. These didn’t require transferring and refilling liquid cryogenic helium every other day to keep these refrigerators cold. However, around 2010, the whole world started switching over to these reliable cryogen-free “dry” dilution refrigerators which suddenly allowed for experiments with superconducting qubits to be done for a lot longer periods of time with no interruption.
Dilution refrigerators are the solution to generate extremely low temperatures of 1 K to several ten mK for developing next-generation technologies. Helium atoms have two types of isotopes: one is common 4He with atomic weight of four and another is the lighter 3He with atomic weight of three. Both helium elements are stable and non radioactive, but their physical properties at cryogenic temperatures are completely different.
A 3He/4He dilution refrigerator is a cryogenic device that provides continuous cooling to temperatures as low as 2 mK, with no moving parts in the low-temperature region. The cooling power is provided by the heat of mixing of the Helium-3 and Helium-4 isotopes.
(1) The refrigeration process uses a mixture of two isotopes of helium: helium-3 and helium-4. When liquid mixture of 3He and 4He is cooled below 0.8 K, mixture separate into two phases: a 3He dense phase (almost pure 3He) and a dilute phase (in which approximately 6％ of 3He is diluted into 4He). The 3He dense phase is lighter, and floats on the dilute phase.
(2) When temperature of the distilling chamber is stabilized to approximately 0.5 K, only 3He will evaporate selectively due to the difference in the vapor pressure (fractional distillation). When 3He is vaporized in the distilling chamber, the concentration difference of 3He in the dilute phase between the distilling chamber and the mixing chamber will be arise and 3He will move upward due to the osmotic pressure from the mixing chamber to the distilling chamber. To compensate for 3He evaporated in the distilling chamber, 3He in the dense phase at the mixing chamber will merge into the dilute phase
(3) When 3He is forced to solve from the dense phase into the dilute phase, refrigeration power will be generate. This process is called dilution, so refrigerators using this process are called dilution refrigerators. The liquid 4He at 0.5 K or lower temperatures has zero viscosity (superfluidity). 3He in the dilute phase can be regarded as a “gas” that can freely move in the superfluid 4He. Meanwhile, 3He in the dense phase can be regarded as a “liquid” state in which interaction of the 3He is strong . Therefore, the dilution process can be regarded as a process in which the liquid state 3He evaporates to gas state 3He. The refrigeration capacity produced in this process is correspond to the latent heat of vaporization.
(4) The 3He evaporated in the distilling chamber is discharged and compressed by an external vacuum pump and returned to the cryostat again. After that, in a conventional dilution refrigerator, 3He gas is cooled down to 1K and liquefied with depressurized liquid 4He (precooling). The liquefied 3He is further cooled through the heat exchanger and returns to the mixing chamber.
(5) As described above, 3He circulates in dilution refrigerators and generates cooling power continuously.
Dilution refrigerators can generate low temperatures from 1 K to several mK, have been widely used for measurement of physical properties, but currently less used in other area. The main problems are the equipment is large, liquid helium needs to be refill periodically, and especially the operation procedures to handle the dilution refrigerators are complicated, which makes it impossible for general users to handle the dilution refrigerators.
Cryogen-free dilution refrigerators
Figure 2 illustrates the schematic diagram of the cryogen-free dilution refrigerator. In the cryogen-free dilution refrigerator, the circulating 3He gas is cooled to approximately 3 K with a mechanical refrigerator (4K-GM refrigerator or 4 K pulse tube refrigerator). Then the 3He gas undergoes Joule-Thomson expansion with the JT valve and liquefies.
The advantages of cryogen-free dilution refrigerators are listed below.
(1) Easy operation (controlled automatically with computers), reduces the physical and mental burdens on researchers and allows easily operation on operators who do not have much experience.
(2) They can be made to be ultra-compact.
(3) It does not consume liquid helium (addressing the helium resource problem).
Meanwhile, the disadvantages are that efficiency is worse than conventional dilution refrigerators and mechanical refrigerators generate vibrations. Conventional dilution refrigerators can cool to minimum temperature of 2 mK, but cryogen-free dilution refrigerators can achieve only 10 mK range
Adiabatic demagnetization refrigerator is another type of refrigerator that can get to 10 mK
An ADR usually gets to about 3K with a helium compressor. That compressor can run all the time, so the refrigerator can sit at 3K indefinitely. To get down to mK temperatures, the ADR works like this:
- Raise the magnetic field surrounding a solid with nuclear spins. This aligns the spins.
- Slowly turn the field off. This allows the spins to randomize their direction, which absorbs entropy from the surroundings and lowers the temperature.
- Once the field is back to zero, we’ve sucked enough heat out of the surroundings to bring them to mK temperatures.
This is all great and it really works, but it’s a “one-shot” process. Once the field is down to zero, you can’t go any lower. Heat from the surroundings, such as the room temperature outer parts of the refrigerator, leak heat into the part you’re trying to keep cold, and since we’ve already lowered the magnetic field to zero, we can’t do anything to remove that heat. Therefore, after cooling the ADR, it starts to warm up (hopefully slowly enough to run your experiment).
It’s typical for an ADR to stay below 100mK for maybe twelve hours, although that number depends a lot on how many wires you have running to the cold part of the ADR. After the temperature rises above what you want, you have to raise the magnetic field again and slowly lower it to re-cool. Raising and lowering the field takes a while and heats up the refrigerator, and that big magnetic field is often incompatible with superconducting qubit experiments, so you can’t run experiments while you’re in that stage of the process.
The dilution refrigerator, on the other hand, runs continuously, so you have as long as you need to run your experiment. That’s a pretty big reason that they’re in common use. Note, however, that other refrigerators aside from the ADR are used in many superconducting qubit labs for tasks where the benefits of a dilution refrigerator aren’t needed and the shorter cold time of an ADR is ok. For example, ADR’s are common for experiments with superconducting resonators, which are used to test the quality of materials that may later be used for a qubit.
IBM’s Goldeneye: the world’s largest dilution refrigerator
To reach this ‘moon landing’ moment, the IBM team developed the largest dilution refrigerator, which will house a future 1,000,000 qubit system. Work is underway to reach the goal of quantum computer capable of surpassing conventional machines by 2023, and this 10-foot-tall and 6-foot-wide “super-fridge” is a key ingredient, capable of reaching temperatures of 15 millikelvin, which is colder than outer space. The fridge gets so cold it takes between 5 and 14 days to cool down.
The very first thought of building something at that scale came from my colleague Pat Gumann while brainstorming long-term, ‘crazy’ ideas in my office in November of 2018, explains Jerry Chow, Director of Quantum Hardware System Development for IBM. At that time, our team was tasked with deploying our first 53-qubit quantum computer in the IBM Quantum Computation Center in Poughkeepsie, NY, a challenge which pushed a few limits in what we could place into a single cryogenic refrigerator at the time. While working on it, it also really made us start thinking beyond, and almost instantly that we will need much larger cryogenic support system to ever cool down between 1,000 to 1 million qubits. This was simply due to the sheer volume required to host, not only all the qubits, but also all of the auxiliary, cryogenic, microwave electronics – cables, filters, attenuators, isolators, amplifiers, etc.
It became very apparent that a new way of thinking in terms of the design would be needed and we started coming up with different form factors for how to effectively construct and cool down a behemoth such as the super-fridge. Some of the challenges we had were purely infrastructural such as how were we going to find a space in the building big enough to start this project and where would we find the capabilities to work with really large pieces of metal.
Some of the most challenging hurdles to overcome includes improving the quality of the underlying qubits, which includes improving the underlying coherence times (the amount of time that qubits stay in a superposition state), the achievable two-qubit gate fidelities, and reducing crosstalk between qubits as we scale up. For that matter, most of these improvements feed into an overall quality measure for the performance of a quantum computer which we have defined called the Quantum Volume. Having a measure such as Quantum Volume allows us to really show progression along a roadmap of improvements, and we have been demonstrating this scaling of Quantum Volume year over year as we make new systems better and better.
The higher the Quantum Volume, the more real-world, complex problems quantum computers can potentially solve. A variety of factors determine Quantum Volume, including the number of qubits, connectivity, and coherence time, plus accounting for gate and measurement errors, device cross talk, and circuit software compiler efficiency.
Our “Goldeneye” super-fridge is very much an ongoing project, which is on target for completion in 2023. It is just one critical part of our long-term roadmap for scaling quantum technology. As we continue to execute on the roadmap we announced in September, we’re pleased to share that we achieved a Quantum Volume of 128 in November and we’re working towards improving the quality of our underlying systems in order to debut our 127-qubit IBM Quantum Eagle processor later this year.
Oxford Instruments NanoScience reckons its Proteox dilution refrigerator will help researchers and start-ups to fast-track the development of next-generation quantum technologies
Oxford Instruments NanoScience is a division of parent group Oxford Instruments, designs and manufactures research tools to support the development, scale-up and commercialization of next-generation quantum technologies. This includes cryogenic systems (operating at temperatures as low as 5 mK) and high-performance magnets that enable researchers to harness the exotic properties of quantum mechanics – entanglement, tunnelling, superposition and the like – to yield practical applications in quantum computing, quantum communications, quantum metrology and quantum imaging.
It’s with this quantum opportunity front-and-centre that the fundamentals of the Proteox dilution refrigerator have been reimagined to support multiple scientific users and a variety of ultra-low-temperature experiments from a single system operating in the mK regime. That scalability is achieved with a side-loading “secondary insert” module that allows samples, communications wiring and signal-conditioning components – basically full experimental set-ups – to be installed and changed whenever necessary.
“Proteox is the largest dilution refrigerator in its class with an extensive capacity for integrating components, experimental services and sample mounting,” explains Harriet van der Vliet, product segment manager for quantum technologies at Oxford Instruments NanoScience. “Modularity and flexibility are key,” she adds, “and we work closely with our customers to offer them tailored solutions and experimental set-ups on standard lead times.”
The University of Glasgow’s quantum circuits group announced in January that it is also using the Proteox dilution refrigerator to support its wide-ranging R&D effort in superconducting quantum technologies, including dedicated initiatives spanning superconducting spintronics, quantum-engineered nanoelectronic circuits and quantum information processing.
“We’re excited to be using Proteox, the latest in cryogen-free refrigeration technology, and to have the system up and running in our lab,” explains Martin Weides, head of Glasgow’s quantum circuits group. “Proteox is designed with quantum scale-up in mind, and through the use of its secondary insert technology, we’re able to easily characterize and develop integrated chips and components for quantum computing applications.”
The University of Glasgow, its subsidiary and commercialization partner, Kelvin Nanotechnology, and Oxford Instruments NanoScience are part of the OQC-led R&D consortium developing specialist foundry and measurement services to support the commercialization of superconducting quantum technologies (see main article). Other consortium partners include quantum computing pioneer SeeQC UK and the SuperFab nanofabrication facility at Royal Holloway, University of London.