EDA Tools For Quantum Chips

Commercially viable quantum computers are at least several years away, but some researchers already are questioning whether existing EDA tools will be sufficient for designing quantum chips and systems. That’s because quantum design requirements at times transcend classical rules about materials, temperature, and structure — rules that are foundational for the majority of EDA products on the market.

This situation could accelerate EDA’s movement toward customization as quantum computing becomes a more attractive business opportunity. McKinsey & Co. estimates the quantum computing market could reach $700 billion as early as 2035. Scientists say quantum computers could drive truly disruptive innovations across a wide range of fields, including pharmaceuticals and AI. But there are a slew of technical hurdles that must be resolved first, especially the production of durable, error-correcting qubits in larger quantities than what is possible today.

Quantum researchers currently use a variety of different materials, techniques, and approaches to develop quantum computers. Because there is no consensus on basic issues like the best approach for creating a qubit, there are likewise different ways of designing the chips. For example, some systems rely on electron spin, while others utilize photon polarization.

Existing EDA tools are generally better suited to the electronic (superconducting) domain than the photonic (optical) one when designing quantum chips, but neither is perfect.

“One of the challenging parts of the superconducting chip design is the complexity of the electromagnetic simulation with the increased number of qubits,” said Mohamed Hassan, quantum solutions planning lead for Keysight Technologies. “Moreover, tuning qubits and resonators and their coupling implies a repetitive cycle of electromagnetic simulation that significantly prolongs the design cycle and increases the computational cost.”

Hassan explained that the design of superconducting qubits composed of Josephson junctions follows a typical microwave circuit design process. “The quantum microwave circuit usually consists of co-planar waveguide resonators coupled to qubits. The resonators are used for two purposes — to read out the state of the qubits and to make them talk to each other. In quantum language, that is entanglement. From a microwave design perspective, the circuit will have many resonance frequencies corresponding to the resonators and the qubits. In fact, qubits can be abstracted by their lumped inductances expected in the quantum regime at very low signal excitation.”

Fig. 1: The top image depicts a four-qubit circuit created in IBM Qiskit Metal and imported to PathWave ADS2023. The square islands represent qubits and the wiggly lines represent entanglement/bus resonators. The bottom image shows EM simulation results using the method of moments. Source: Keysight

While this has some of the elements of existing semiconductor design, there is a whole different set of optimization techniques required in the quantum world, some of which are still evolving. As a result, there are no agreed-upon best practices and methodologies, and it’s not clear when there will be. However, both will be essential for widespread commercial adoption.

“The chip design entails engineering the resonators and the qubits to land the different frequencies of the circuit to targeted values, typically in the few-GHz regime, and adjusting the coupling between the resonators and qubits to meet certain quantum parameters,” said Hassan. “When the qubit changes its state, we will be able to see that on the microwave signal response of its associated readout resonator with a shift in the resonance frequency.”

Planar microwave circuits lend themselves to a very cost-effective electromagnetic simulation solution using the method of moments, which instead of solving for the electric field in the full volume, solves only for the currents on the metal surface. That approach significantly reduces the computational cost. Quantum planar microwave circuits can leverage the same benefits, enabling faster design cycles for large quantum circuits. But scaling and accuracy requirements are more demanding for quantum circuits, prompting the need for innovative solution techniques and skilled engineers to meet the new challenge.

Using optical design approaches requires development of nonlinear shapes for photons — curved waveguides rather than an electrical junction, so that optical signals can move smoothly through a chip. “[These are] very large curvilinear structures that not only have to be very accurately shaped circles and waveguides, but also very closely coupled to other waveguides when you make a photonic interconnect in the quantum regime or even in the room temperature regime,” said Ted Letavic, corporate fellow and senior vice president of technology and innovation at GlobalFoundries. He noted that EDA tools generally are sufficient to design waveguides and coupling structures for photonic receivers and data center communications. “However, with the uniqueness of the quantum requirements, substantial updates are needed to the EDA tools in the way we handle non-rectangular shapes.”

Other important changes will be required to reflect the nonlinear optical properties of the materials involved in the photonic domain, which today includes barium titanate, strontium titanate, molybdenum, and silicon alloys.

In contrast, quantum chip design in the electronic domain is mostly about “optimization,” according to Letavic. “It’s very similar to other electronic systems,” he said. “There are normal EDA layout rules for some ‘keep out’ zones that you have to pay attention to, depending on noise. We have to be aware of radio frequency (RF) shielding and that sort of thing. But the techniques are well known. I wouldn’t say there are any major showstoppers there.”

That changes when one looks beyond chips to consider designing packaging simulation and the impact of temperature. “If you have a quantum computer that has three or four different temperature regimes, going from 4 millikelvin (-273°C) all the way up to room temperature, the stresses and strains on the packaging are rather extreme,” he said. “There is a tremendous amount of work that still has to be done to quantify the reliability and ruggedness of the various packaging solutions that are necessary to implement the next generation of quantum computers.”

The power required to keep quantum systems cool could lead to a chiplet model, which also has implications for EDA. “What has to happen in these quantum systems is a chiplet-based format,” he said. “Some of those chiplets are going to be at very low temperatures and some of them are going to be at room temperature. The solution will be chiplet segmentation. The EDA tools have to be able simulate a single chip in the system very accurately and with high fidelity, and they also have to very accurately model the interconnects going to chips and other parts of the system at different temperatures.”

To be sure, not everyone in the quantum world questions the adequacy of the existing EDA ecosystem. William McGann, CTO and COO of Quantum Computing Inc., said the company recently purchased a license for an existing EDA product that he believes will provide more than what is needed to build out the company’s chip designs. McGann says the qubits are photonic, and the design makes use of lithium niobate as well as certain challenges associated with translating the technology to a mass-produced chip.

“When you do smart cutting, there’s a certain amount of damage based on the energy and the cleavage plane that you cut the crystal across,” said McGann. “How do you polish those defects out of the surface? How do you measure the active volume of your device when building photonic bandgap materials? Certainly, the purity is very important from an optical perspective. Some of the metrology challenges, such as how we’re going to measure what we know to be good, can probably advance some of the laser technology that we’ll need beyond what we have today. Most of the processes we anticipate using have already been demonstrated publicly, so now it’s really about us building our recipes.”

Still, GlobalFoundries’ Letavic predicts considerations like the ones he described could compel quantum researchers to seek out custom solutions. That could amplify the existing EDA trends toward increasing customization and bespoke silicon. “The EDA tool environment often only focuses on the hardware portion of the ecosystem, but what’s really going to determine whether these quantum computer architectures live up to the promise of quantum is the rest of the stack,” said Letavic. “There is going to be so much difference in software and algorithms and quantum sources that I don’t see this narrowing down to an off-the-shelf solution over the next decade.”

He predicts these EDA solutions will be extremely customized, beginning with the properties of the qubit source itself — photonic qubits with their nonlinear properties and electronic qubits with their extreme temperature requirements.

John Ferguson, director of product management at Siemens Digital Industries Software, said much of the design flow in quantum computing is similar to classical computing. “It’s like the EDA of the mid-to-late 1980s. The automation is not there. What is forcing the industry to innovate are the issues with curvilinear shapes, which have long been difficult to design with existing tools. We’ve known it was coming for a long time, but everyone said ‘Let’s wait until there’s really a need.’ Now we can’t avoid it anymore and everybody’s working on it.”

Traditional EDA tools for design rule checking do not lend themselves well to bent shapes, however, because in the IC world wires and transistors are drawn as rectangles.

“But increasingly, things of a curved nature are becoming a fact of life, not just in quantum computing, but in photonics and other areas. There are proposals going in to change the Oasis format so that we can represent things that are curvilinear,” Ferguson said. “Quantum design is currently a bit more of a mix of art and science, with a little bit of intuition thrown in. It’s more like doing a custom analog block where you know exactly what you want, you know how they’re supposed to interact with each other, and you’re moving them graphically, then running some form of simulation. It’s a lot of what I call ‘construction by correction.’ Developing EDA tools to meet the needs of quantum computing researchers is an important and difficult problem. And it is far less difficult than some of the technical challenges around scaling quantum computers, such as developing more of the right type of qubits. It’s an interesting challenge, but it’s surmountable.”