A hole punched in a card, or the lack thereof; a switch turned on or off; the numbers zero or one. These can all represent the binary digit, or ‘bit’: the most basic unit of information in today’s digital world. Using long strings of varying bits, the electronic chips in our smartphones compactly store entire libraries and instantly make calculations that take hours by hand.
However, a very different paradigm of computing is emerging: one built on the quantum bit, or ‘qubit’. Rather than the electric signals of bits, which can only represent one of two states, qubits are encoded into quantum states of subatomic systems, such as photons or the electronic states of atoms. This allows a qubit to have multiple possible states: a versatility that’s foundational for quantum computers built on radically different principles from ‘classical’ machines widely used today.
“Quantum computing ushers in a new paradigm, alongside classical computing, that can solve a wide range of computational problems significantly faster, such as those encountered in chemistry and drug discovery, as well as in logistics and financial optimisation,” said Keng Hui Lim, Assistant Chief Executive of A*STAR’s Science and Engineering Research Council.
At A*STAR, research institutes like the Institute of High Performance Computing (IHPC), the Institute of Materials Research and Engineering (IMRE) and the Institute of Microelectronics (IME) are working not only to develop the foundational tools needed to make quantum computers a practical reality, but also to identify their advantages over classical counterparts.
“In partnership with Singapore’s quantum research ecosystem, A*STAR’s research takes a holistic approach to develop end-to-end, full-stack software solutions for quantum algorithms and applications, as well as quantum materials and devices to build scalable quantum computers,” said Lim.
A national strategic focus
As part of that approach, A*STAR collaborates with national initiatives that drive quantum computing research, such as the National Quantum Computing Hub (NQCH). A joint initiative between IHPC, the Centre for Quantum Technologies (CQT) and the National Supercomputing Centre (NSCC) Singapore, NQCH supports the development of quantum computing software, middleware and hardware. The platform also works with industry to explore quantum solutions for real-world challenges, and hosts programmes for talent development in an emerging field.
Another key initiative is the National Quantum Federated Foundry (NQFF), which builds and provides access to capabilities needed to design, fabricate and characterise quantum devices. To support local research needs in quantum hardware, NQFF taps into and augments an existing network of cleanrooms across Singapore. The platform draws on IMRE’s state-of-the-art device design and characterisation facilities, as well as device fabrication facilities at IME, the National University of Singapore (NUS) and Nanyang Technological University (NTU).
Both NQCH and NQFF are supported by the Quantum Engineering Programme (QEP), which aims to establish a competitive quantum engineering research community in Singapore.
Simulations in the quantum code
At NQCH, Jayne Thompson, Group Manager of IHPC’s Quantum Algorithms and Physics team, delves into quantum software by exploring how quantum algorithms can model and simulate complex, real-world physical systems.
“When we design quantum algorithms, they should harness the properties of quantum systems to do computations with less resources than any possible classical method,” said Thompson. “This can include completing problems faster, using far less working memory or input data, or delivering more accurate heuristics than what is feasible on classical machines.”
These advantages are possible due to a quantum phenomenon known as superposition, where a qubit is in multiple states (different configurations) at the same time.
“A quantum computer reaches the final answer faster by passing through many states in superposition,” said Thompson. “Through careful engineering, we harness the quantum interference between those states to amplify the desired solution.”
Superposition makes quantum computers a powerful computational tool to model and simulate classical systems like time-series events, where “we can simulate many possible futures,” said Thompson. “In tandem with other quantum algorithms, this can speed up time-series analysis tasks—for example, risk analysis for tail events, which occur rarely but make huge impacts.”
Such tail events include black swan-type events which can send financial markets crashing. One project by Thompson’s group involves analysing S&P 500 data to predict the probability of rare events, and quantify the risk they pose to investment portfolios. Another includes computational fluid dynamics problems used to describe traffic networks and models of epidemic spread in populations.
Quantum tools for quantum problems
As quantum computers are built on quantum mechanics, they also have a natural advantage when simulating quantum systems, such as chemical reactions or drug-protein interactions at the nanoscale.
“To simulate a quantum system on a classical computer, we need to use extremely high-dimensional data,” said Thompson. “That storage quickly becomes prohibitively costly; it’s also almost impossible to update those stored states under the simulation model. These scaling issues rapidly force researchers to make approximations when modelling these systems with classical architectures.”
Conversely, the qubits of quantum computers can more compactly describe those states for digital simulation; by modelling atomic-scale behaviour with atomic-scale hardware, computations can be easily scaled up to classically intractable levels.
Quantum chemistry is a focus area for IHPC Principal Scientist Adrian Mak, whose team has been designing quantum algorithms for the field with CQT Principal Investigator Dimitris Angelakis and colleagues.
“By understanding the wave-like properties of electrons and how they affect atoms and molecules, we can better understand the electronic structures of atomic and molecular systems,” said Mak. “This allows us to predict their geometries, spectra and reaction mechanisms, which then helps us to design new chemical reactions and materials.”
Mak, Angelakis and colleagues recently produced an algorithm that calculates correlation energy—the measure of how much an electron’s movement is affected by other electrons in the same quantum system—using quantum circuits with two-qubit gate depths. Capable of scaling linearly as more qubits are added, the team has successfully demonstrated the algorithm’s use in cloud-accessed quantum computers.
“We’ve also developed a quantum algorithm that can encode a multi-electron state using quantum circuits with polylogarithmic scaling of depth, with respect to system size,” said Mak. “Both these projects were supported by QEP 2.0.”
NQCH efforts in this area led to the release of Qibochem, a plugin for quantum computational chemistry that works with NQCH’s open-source quantum computing platform, Qibo. “Qibochem avails quantum computing for chemistry to the public without a need to intimately understand all the quantum details involved,” said Mak.
Building the hardware
Each qubit in a quantum computer presents a complex manufacturing challenge. To encode information in a single nano-sized particle, how do you keep it stable, measurable and free from interfering forces?
Working with NQFF, IME Principal Scientist Hongyu Li’s group aims to fabricate hardware that meets requirements posed by different quantum computing architectures, such as trapped ion and superconducting qubits. The team is focused on packaging and process integration: designing functional quantum chips and fleshing out the steps to fabricate them.
“To understand the superconducting testing process, we worked closely with IMRE’s cryogenic testing lab and CQT testing labs based at NUS and NTU,” said Li. “IMRE also provided support in characterising materials and developing processes for film deposition.”
From 2017 to 2021, Li’s group worked with IMRE’s Quantum Technologies for Engineering (QTE) department to develop the fabrication process and packaging approach for a strontium ion (Sr+)-based 3D surface ion trap design that can hold over 200 ions.
“Since 2022, we’ve been developing a barium ion (Ba+)-based ion trap with NQFF Director Manas Mukheerjee in a QEP-supported project,” said Li. “As Ba+ ion trap chips need high voltage RF—over 300 volts—this project needs multiple wavelength waveguides, as well as grating coupler integration under metal plates. Other ion trap chip integration processes are also in development.”
Li’s team is also creating hardware that can function at the very low temperatures needed for stable qubits. “With CQT Principal Investigator Rainer Dumke, we’re developing cryogenic through-silicon via (TSV) applications within resonators and RF characterisation; as well as approaches to confine top quit chips to cryogenic interposer distances,” said Li.
In another project under A*STAR’s DELTA-Q Strategic Research Programme, Li‘s group and that of IMRE Senior Principal Scientist Johnson Goh developed a cryogenic interposer that copes with different transition temperatures.
Work by Li’s group has attracted attention from the quantum hardware industry. “We recently submitted a joint proposal to an Australian startup on cryogenic packaging, and are in discussion with a Europe-based company for a project using our cryogenic interposer designs,” said Li.
The best of both worlds
With quantum computers well-suited to weighty computations and classical computers likewise to hefty data loads, a promising research direction involves pairing both architectures together.
“Hybrid computing systems leverage the strengths of classical high-performance computing (HPC) for general tasks such as big data processing, and quantum computing for highly complex and intensive computations,” said Yi Su, IHPC Executive Director and NQCH Co-Principal Investigator. “This approach holds great promise for problems that pose both ‘big data’ and ‘big compute’ challenges.”
At NQCH, IHPC Senior Principal Scientist Hoong Chuin Lau and his research group are working with IBM and Japanese universities such as Tokyo University, Tokyo Institute of Technology and Tsukuba University to identify useful hybrid classical-quantum techniques in logistics and transportation.
“The logistics sector collects a huge amount of data from orders and deliveries: locations, times, volumes and types of goods, order status, routes used, and so on,” said Lau, also a Professor of Computer Science at Singapore Management University. “The challenge is deriving good predictions—like estimated delivery times—from all this data, and developing good stochastic optimisation models for tasks like order despatch and route planning.”
While classical HPC can handle big data, it often struggles with using it to compute optimised solutions to logistics problems, such as the most efficient routes for a fleet of delivery vehicles, or the best combination of goods to load in a container, especially when their inputs are uncertain. Lau and colleagues are examining how hybrid quantum algorithms like VQE and QAE can help bridge the computational gaps.
Hybrid computing is also of interest to researchers like Chandra Verma, Senior Principal Investigator at A*STAR’s Bioinformatics Institute (BII), who see their potential to support AI-powered drug discovery.
“We’ve already seen the first example of a quantum machine learning (QML) framework for small molecules that outperforms its classical counterpart,” said Verma. “There’s potential to develop hybrid QML models that achieve low generalisation errors, even with limited training data available—as is the case with small molecule libraries.”
BII has recently begun to explore quantum computing applications in drug discovery problems such as protein conformations, peptide design and RNA folding.
“We’re also looking at quantum-enhanced, artificial intelligence (AI)-powered assays to monitor T cell response in infectious disease control and immunotherapy, as well as quantum-deep learning models that enhance immunotherapy response prediction,” added Sebastian Maurer-Stroh, BII Executive Director.
Future prospects: other research perspectives
“Quantum computers could help combine various sources of health data—genomics, proteomics, clinical records—by processing them in parallel and modelling their complex statistical relationships. This could help speed up drug response predictions and molecular diagnosis of complex diseases.”
—Dennis Wang, Senior Principal Investigator and Bioinformatics Platform Head,
A*STAR Institute for Human Development and Potential (A*STAR IHDP)
“These systems may have synergistic effects with AI. For instance, the data used to train AI models are often probabilistic and costly to evaluate; in such settings, quantum computing-based simulations may provide improved estimates with fewer evaluations. Moreover, quantum circuits are similar to deep neural networks in that they can approximate high-dimensional functions from large data sets, but a promising advantage is that they often need fewer parameters.”
—David Bossens, Senior Scientist, Centre for Frontier AI Research (CFAR)
“With the Port of Singapore aiming to deal with 65 million twenty-foot equivalent units (TEUs) of cargo annually in future, optimisation algorithms on classical computing facilities may not be able to handle the complex optimisation problems involved with so many dynamic factors and players in making sure port operations are at their most efficient and sustainable. Quantum computing may enhance AI-based solutions by speeding up optimisation and parameter convergence in this area.”
—Xiuju Fu, Senior Principal Scientist and Director (Maritime AI Research Programme),
Institute of High Performance Computing (IHPC)
A practical quantum horizon
While research communities worldwide are racing to bring the first scalable and practical quantum computers beyond the lab, it remains important to consider the technology’s potential strengths without introducing a sense of hype, said Ping Koy Lam, A*STAR Chief Quantum Scientist.
“Quantum computing is still a very nascent research area; currently, quantum computers are aimed at deep problems that require a lot of computing time on classical computers,” said Lam. “As quantum computers are expected to have very small input bandwidths for the foreseeable future, it’s hard to predict the roles they might play in big data problems.”
“At present, quantum computing’s strengths lie in ‘big compute, small data’; problems with relatively small data sizes, but very high computational complexity,” added Yi Su. “For large-scale data processing tasks, classical computing remains more practical and efficient with current technologies. However, hybrid computing offers opportunities to target ‘big data, big compute’ problems in future.”
To continue exploring quantum’s technological capabilities, A*STAR continues to support research efforts across multiple fields in Singapore’s quantum ecosystem. In 2023, A*STAR launched the Quantum Innovation Centre (Q.InC) to advance use-inspired basic research within the agency’s scientific community and to focus on translational research for real-world impact. Powered by expertise and facilities from IHPC, IME, IMRE and the National Metrology Centre (NMC), Q.InC aims to develop next-generation translational quantum technologies, as well as talents in quantum science and engineering, supplementing NQCH and NQFF efforts.
Beyond pursuing an advantage over classical systems, Q.InC is also exploring novel quantum hardware applications in quantum photonics, quantum materials and quantum sensing.
“A*STAR plans to work with the research ecosystem including university and industry partners to push the state of the art, and develop breakthrough solutions for real-world use cases,” said Keng Hui Lim. “Such collaborations are necessary to advance quantum computing research, and contribute to Singapore’s National Quantum Strategy.”
Other National Quantum Strategy programmes with A*STAR support include the National Quantum Processor Initiative to build local capacities for quantum processor design and fabrication; as well as the National Quantum Software Programme to support research in positioning, navigation and timing; biomedical sensing and imaging; and remote sensing applications.
