Intel is placing two research bets related to Quantum computing: creating a superconducting quantum test chip, and researching "spin qubits," an alternative structure, which draws on the company's expertise manufacturing silicon transistors.
Quantum computing is heralded for its potential to tackle problems that today's conventional computers can't handle. Scientists and industries are looking to quantum computing to speed advancements in areas like chemistry or drug development, financial modeling, and even climate forecasting.
While there's been significant progress, quantum computing research is still nascent. Many problems must be solved and many architectural decisions must be made. For example, it's not yet clear what form quantum processors (or "qubits") will take. That's why Intel is placing two major research bets and investing in them equally.
One possible form is superconducting qubits. Intel is making fast progress in developing this type of test chip, which others in the industry and academia are also pursuing. At last January's 2018 Consumer Electronics Show in Las Vegas, Intel unveiled "Tangle Lake," a 49-qubit superconducting quantum test chip. Achieving a 49-qubit test chip is an important milestone because it will allow researchers to assess and improve error correction techniques and simulate computational problems.
Intel is also researching another type called spin qubits in silicon. Spin qubits could have a scaling advantage because they are much smaller than superconducting qubits. Spin qubits resemble a single electron transistor, which is similar in many ways to conventional transistors and potentially able to be manufactured with comparable processes. In fact, Intel has already invented a spin qubit fabrication flow on its 300mm process technology.
What is a spin qubit?
Spin qubits highly resemble the semiconductor electronics and transistors as we know them today. They deliver their quantum power by leveraging the spin of a single electron on a silicon device and controlling the movement with tiny, microwave pulses.
Electrons can spin in different directions. When the electron spins up, the data signifies the binary value 1. When the electron spins down, the data signifies the binary value 0. But, similar to how superconducting qubits operate, these electrons can also exist in a "superposition," which means they have the probability of a spin that's up and down at the same time and, in doing so, they can theoretically process tremendous sets of data in parallel, much faster than a classical computer.
Among the challenges researchers must overcome before quantum computing can become a commercial reality, is the incredibly fragile nature of qubits. Any noise or unintended observation of them can cause data loss. This fragility requires them to operate at extremely cold temperatures, which creates challenges for the material design of the chips themselves and the control electronics necessary to make them work. Superconducting qubits are quite large and they operate in systems the size of 55-gallon drums, which makes it hard to scale up the design of the quantum system to the millions of qubits necessary to create a truly useful commercial system.
Spin qubits, in comparison to their superconducting counterparts, offer a few advantages in addressing these challenges.
They're much smaller in physical size and their coherence time is expected to be longer - an advantage as researchers aim to scale the system to the millions of qubits that will be required for a commercial system.
Silicon spin qubits can also operate at higher temperatures than superconducting qubits (1 kelvin as opposed to 20 millikelvin). This could drastically reduce the complexity of the system required to operate the chips by allowing the integration of control electronics much closer to the processor. Intel is working with QuTech to explore higher temperature operation of spin qubits with interesting results up to 1K (or 50x warmer) than superconducting qubits. The team is planning to share the results at the American Physical Society (APS) meeting in March.
According to Intel, the design of the spin qubit processors highly resembles the traditional silicon transistor technologies. While there are key scientific and engineering challenges remaining to scale this technology, Intel has the equipment and infrastructure from decades of fabricating transistors at scale.
This week at the American Association for the Advancement of Science (AAAS) Annual Meeting, QuTech, will present on its success creating a two-qubit spin-based quantum computer that can be programmed to perform two simple quantum algorithms. This development paves the way to larger spin-based processors capable of more complex applications.
In recent years, scientists have achieved ever better qubits. These few qubits are now measured and manipulated with such high reliability that useful programming becomes important. To be able to perform universal quantum calculations, operations that control the state of individual qubits and intertwine multiple qubits in a controlled way are required. This set of operations must be available in such a way that we can place them in any order to perform different algorithms - this is what the researchers call programmability.
The scientists from QuTech and the Kavli Institute of Nanoscience Delft, and their collaborators from the University of Wisconsin, focus on qubits in silicon chips. These qubits are created by 'dialing in' a few electrons using controlled electrical signals. Silicon is a promising material, as it is fully developed by the current computer industry and naturally causes little disruption to our qubits.
Using the reliable operations in the two-qubit quantum chip, the scientists managed to program and execute the first quantum algorithms. The first algorithm, the Deutsch-Josza algorithm, tests whether a function is 'even' or 'odd'. It's like you only have to throw a coin once to see if both sides are the same or not. They also performed the so-called 'Grover-search algorithm'. This algorithm searches for the right answer in one go in an unsorted set.
This example of a programmable quantum chip with reliable silicon qubits is an important milestone towards realizing reliable and scalable quantum calculations. Scientists showed that the qubits are controllable, can be entangled with high reliability and that operations can be combined to perform a quantum algorithm.
In order to be able to execute more complex algorithms, the researchers will develop silicon quantum chips with more qubits, both in the Delft cleanrooms and in industrial cleanrooms with Intel.
Intel has invented a spin qubit fabrication flow on its 300 mm process technology using isotopically pure wafers sourced specifically for the production of spin-qubit test chips. Fabricated in the same facility as Intel's transistor technologies, Intel is now testing the initial wafers. Within a couple of months, Intel expects to be producing many wafers per week, each with thousands of small qubit arrays.
Going forward, Intel and QuTech will continue research on both superconducting and spin qubits across the entire quantum system - or "stack" - from qubit devices to the hardware and software architecture required to control these devices as well as quantum applications.