IBM’s quantum computing system is powered by a chip no larger than a quarter. These machines can solve complex problems even the best classical computers cannot. This is only one piece of the puzzle. The computing infrastructure supporting the work of quantum chips is like a Russian Doll, with complex interconnections in a Rube Goldberg-like device.
With all its complexity and design, a quantum computer is still a machine that performs operations using both hardware and software. Some of these actions are similar to those performed on classical computers. You’re curious to know how they work? Popular Science guided the IBM Yorktown Heights campus’ quantum center. Look inside the quantum center, starting with a qubit (more about that in a minute) and moving out bit by bit.
It’s cold
Objects to exhibit quantum properties must either be small or cold. This layered chandelier, which looks upside down like a gold steampunk cake, is known as a dilution fridge by IBM. IBM created this infrastructure for its 50-qubit chip. It keeps qubits stable and calm. The fragment comprises multiple plates, which get colder as they move closer to the ground. Each plate has a different temperature. The topmost layer is at room temperature.
More tubes feeding into the lower levels introduce another closed cycle of cryogenic material comprising a mix of helium isotopes. The lower levels are provided with more lines that present a closed-cycle cryogenic material made of helium mixtures.
The hidden support infrastructure of the chandelier is located in the rear of the housing structure. The cryogenic infrastructure includes the gas handling system, pumps, and temperature sensors. Then there are the classical control electronics, which were custom-built. IBM’s quantum cloud service allows users to orchestrate a series of gates and circuits. These are then converted into microwave pulses aligned and sequenced correctly before being distributed into the system to control the qubits. The readout pulses are used to retrieve the qubit states, which are then translated into binary values for the users.
Qubits, an artificial atom, and a ‘quantum molecule.’
In classical computers, information is represented by binary bits of one or zero. Qubits represent quantum information. These can be a combination of 0 and 1. Superposition is the term used to describe this phenomenon. Superposition is a common phenomenon in the real world. Zaira Nazarios, IBM Quantum’s technical lead for theory, algorithms, and applications, says music is a superposition. It’s a Waveform and therefore has an amplitude between zero and one. It has a phase and, like all waves, can interfere.
The qubits are superconducting and have been packaged in a way that looks like a printed board. The circuit board has wires and a coaxial cable for input and outputting signals. IBM is working on more compact solutions for newer models with a higher-qubit chip. This involves wiring and integrated components. The components will be more excellent if there is less clutter. It takes around 48 hours to cool down a computer quantum to the desired temperature.
To ensure that the quantum computer functions correctly, it is essential to thermally shield and isolates each plate to stop black body radiation from damaging it. Engineers vacuum seal the device to prevent unwanted photons, electromagnetic radiation, and magnetic fields.
Qubits can be controlled by microwave signals ranging from 4 to 7 gigahertz. Classical electronics uses microwave pulses to transmit input and output signals down cables. As the signal travels along the chandelier, it passes through filters, amplifiers, and attenuators.
IBM uses a lot of superconducting quantum bits. These tiny pieces of metal sitting on the wafer are used to create the chip. Metal is composed of superconducting metals like niobium and aluminum. Josephson Junction is made by layering an extremely thin insulator between two superconducting materials. This provides the nonlinear element required to convert a superconducting system into a qubit.
Jerry Chow is the director of IBM’s quantum infrastructure. He says, “We are building quantum examples of oscillators.” Oscillators are devices that convert the direct current (in this instance, microwave photons), from a power source, into an alternate wind or wave.
Chow says that a nonlinear oscillator differs from harmonic oscillators because it has an uneven spacing of energy levels. Chow says that once you have this, you can isolate two of the lowest levels to act as quantum zero and one.
Imagine a hydrogen atom. It has different energy levels from a physics perspective. The wavelengths of the light that hit this atom can promote it into other states. The qubit performs a similar function when microwaves are aimed at it. Chow explains, “You have an artificial atom.” We have a quantum of energy that we can move by using the right amount of microwave photons at a specific pulse for a particular duration. This will either excite or re-excite this quantum of energy in this nonlinear oscillator.
In a classic computer, the on-state is one, and the off-state is zero. In a quantum computing system, the zero states is the ground state of the artificial atom. A pulse of microwave energy will excite the qubit, promoting it from zero to one. The qubit would return to the ground state if it were hit with this pulse again. Chow explains that if you reduce the energy or time by half, it would be possible to drive a superposition. If you measure the state of a qubit using a resonator, you have a 50% chance that it is in zero and a 50% chance that it is in one.
Users can experiment with circuit elements, pulse frequency, duration, and energy to couple, swap, or perform conditional operations like building an entangled state and combining single-qubit functions for universal computation on the entire device. Waves can be amplified or deconstructed when they intersect.
What is the use of qubits?
Over the past couple of years, quantum computers’ practical applications have changed. Katie Pizzolato is the director of strategy research and applications at IBM Quantum. She says, “If I look at what was being done with the system during that time frame, 2016, 2017, 2018, it was using quantum for quantum research… condensed material physics, particle mechanics, things like this.” The key to this will be making classical resources quantum-aware. We need to help people who are experts on their subject understand how to use quantum but have yet to become quantum experts.
IBM’s interest in quantum problems can be divided into three categories: chemistry, materials, machine learning, and optimization (finding a solution from a list of options). It is essential to not use a quantum computing system for every aspect of the problem but only on the most challenging parts.
