By Andre Saraiva, UNSW
As the COVID-related restrictions are lifted in Australia, 800 people, among quantum entrepreneurs, scientists, government and defence representatives, and venture capitalists, gathered this week both online and in person at Doltone House, by the Sydney Bay. For many, two years have passed since they were in person for a conference. These two years have been transformational in the Quantum Industry, and everyone was keen to showcase their successes and views for the next decade.
Recognising the importance that Australia has as a country in the world scene of quantum technologies, the organisers led by the Sydney Quantum Academy attracted top officials from government and global companies like Google, IBM and Microsoft. Xanadu, the platinum sponsor of the event, was joined by a long list of other startups in the field such as PsiQuantum, Q-CTRL, Alpine Quantum Technologies, Silicon Quantum Computing, Quintessence Labs, Quantum Brilliance, and so on.
Here are some highlights from the three days of conference.
Day 1: After opening remarks by Peter Turner (CEO of Sydney Quantum Academy), Bernie Hobbs (celebrity science communicator and event MC) and the New South Wales Minister for Science, Innovation and Technology, The Honorable Alister Henskens, started a day consisting mostly of panels with a few keynote speakers.
In the first panel called Quantum State of the Nation, a review of the exciting outcomes of the Australian ecosystem was provided by the panelists. The status of Australia in the world context was discussed along with some controversial discussions raised by the public regarding protectionist measures and discussions about export control and its impact in quantum business and the growing concern regarding the inhomogeneity of IP rules across Universities in Australia.
Next, a keynote presentation was made by Dr. Cathy Foley, Australia’s Chief Scientist and herself a veteran in quantum innovation (Dr Foley invented and commercialised sensors based on high temperature superconductors for mining applications). She pointed out how Australia has a dichotomy between being strong in all layers of the quantum computing and networking stack, but also struggling with lack of focus and resulting dilution of investments.
After coffee, quantum software theorists gathered to discuss one of the hottest topics. This was the first time in the conference where a clear shift in narrative could be observed. Very limited emphasis on NISQ applications and a much stronger argumentation in favour of fault tolerant quantum computers could be seen. With multiple examples, the panel made the case that despite some hype in the field, all truly relevant applications of quantum computing (including chemistry) require resources comparable to those set by Shor’s algorithm.
Simone Severini, Director of Amazon Web Services (AWS), highlighted the continued relationship between Amazon in general and Australia, and emphasized that this relationship has already seen some fruits in its quantum arm, AWS Braket.
On the topic of quantum sensing, Professor John Close from Australian National University kicked off with an excellent description of where is the true quantum advantage in sensing and where is quantum mechanics just incidental and not necessarily a technological advantage. Quantum sensing could be considered one of the first applications to achieve quantum advantage – for instance SQUID magnetometers have been a reality for decades. However, the new applications being pursued try to leverage quantum mechanics not only for the functioning of the devices, but also to use coherence and entanglement to beat the limits of shot noise in uncorrelated repeated measurements. In general, there was consensus that often a quantum sensor would work in conjunction with classical counterparts to complement each other in terms of accuracy, drift, dynamical range and so on.
The panel on “Preparing business for a quantum future” showed the many aspects that might drive the decision for different enterprises to adopt quantum technologies earlier or later. Companies that offer professional services that include cybersecurity, such as giants EY and KPMG, are starting to campaign for quantum readiness in Australia amongst companies that require a high level of security. An example that was mentioned was that of organisations that are already collecting encrypted data for the prospect of decrypting it based on a future quantum computer with the capacity to run Shor’s algorithm. Other companies, such as Japanese giant NEC represented by veteran David Garvin (former Rigetti, QxBranch, and financial services sector) are taking a cautious position in not hyping the potential applications of quantum computers. Perhaps in the opposite side of the spectrum, Peter Johnson from Zapata Computing made the case for going out and trying quantum algorithms instead of waiting for formal mathematical proofs of quantum advantage.
Still on the same panel, Xanadu’s CEO Christian Weedbrook revealed interesting information about his company’s plans. Less emphasis was put towards their Pennylane software platform and more on their new focus on fault tolerant quantum computing. Xanadu, which was originally focused on NISQ applications of bosonic photon sampling with programable photonic circuits, seems to be now setting sail to a grander long-term roadmap. This follows their release of a 38-page-long blueprint last year, five years after the company incorporation. While this prepares their technology for any potential application of quantum computing, Weedbrook says they will continue to focus on chemistry applications as a business strategy.
The last part of the day was focused on scaling to qubits by the millions. First, David Reilly from Microsoft fed the public a spoonful of reality, discussing the realistic challenges of simultaneous operation of millions of qubits, interfacing to classical controllers, heat loads and wiring difficulties. He went as far as saying he does not see how a quantum computing architecture will be able to scale if not by using semiconductors and RF control, with limited space for multi-chip integration, transduction and microwave signals.
Then a panel of experts in the leading qubit platforms for scalability discussed the “Race to 1 million qubits”. Rainer Blatt, a legend of the scientific world and now an entrepreneur through Alpine Quantum Technologies, made the point that with all-to-all connectivity the prospects for low overhead in error correction are excellent. He said he can imagine his ion-trap technology scaling to a thousand qubits, but has a hard time imagining a million-qubit architecture. On the opposite side of the spectrum, Professor Andrew Dzurak raised the point that as processors grow, the error rates in physical qubits become harder to improve, and that we should realistically consider that relevant quantum applications will need closer to a billion qubits.
Day 2: The second day started with a refreshing sight: Jay Gambetta from IBM flew into Australia, just days after the borders opened for the general public, and decided to give his talk in person. His focus was less on the progress within IBM in their hardware and more on the strategies to create a global community of users and developers.
The next talk was from Dylan Saunders, Lead Quantum System Architect at PsiQuantum. He raved about the significance of the series of papers that brought to life the idea of Fusion-Based Quantum Computation, and how that could be a path forward for other technologies besides photonic qubits. He also explained the overall architecture, including controllers and sensors, called Q1. Finally, he showed results of their engagement with GlobalFoundries and how it led to significant technical improvements in their single-photon detection system.
Hartmut Neven, Engineering VP of Google, gave a surprisingly technically focused presentation that included an excellent explanation of how logical qubits exponentially supress errors in quantum computation and the general roadmap that all companies seeking fault tolerance need to follow. He then discussed the error budgets in their hardware system. The second half of his talk was focused on software applications, using as an example the possibility to perform quantum machine learning in quantum data, such as the analysis of potentially entangled photons coming from astronomical sources or man-made satellites.
A very interesting panel on “Responsible innovation in quantum technology” followed, where some important issues in quantum computing business and applications were raised. For instance, Tara Roberson from EQUS raised important points about the use of machine learning in “black-box decision making” and how it creates unexplainable criteria in setting decisions that may alter people’s lives. These are potential exacerbations of current problems faced by AI applications, which might be significantly increased with quantum computational power. The NSW Chief Data Scientist Ian Oppermann stressed that this is an important element in the adoption of AI, but it has to be balanced with the potential gains that the population might have in better informed policies, such as the use of AI for identifying vulnerable citizens. Perhaps a more immediate ethical issue was raised by Scott Aaronson from the University of Texas at Austin, a renowned advocate for sobriety in “quantum claims”. He says he is appalled by how quickly people have moved from an academic stance of peer reviewing and scientific rigour to succumbing to the temptation of hype and inflated assertions.
Quantum networking and the prospect for an Australian national network was analysed by a panel of experts. Vikram Sharma, a true veteran in the quantum industry, has steered Quintessence Labs since 2008 and therefore could give a very solid understanding of the challenges and opportunities for quantum key distribution and true random number generation. A topic that was addressed by all panel members relates to the geopolitical implications of a multinational network as well, and how low-orbit satellites might be fast enough to convey quantum information stored in long-lived quantum memories, thus avoiding the necessity to have cables crossing areas that might be subject to adversarial attacks and sabotage. Peter Rohde from UTS also underscored that a quantum network not only is the means for quantum communication, but could potentially integrate quantum computers in different locations. This could multiply their power since shared entanglement may create the means for exploring the exponential growth of computational power in qubit systems.
Michel Biercuk from Q-CTRL showed the successes that his startup has had in improving qubit quality and, in general, preparing hardware makers for a scalable strategy for quantum control.
Then, Michelle Simmons from Silicon Quantum Computing discussed the prospects of building a quantum computer in Australia. She is an important voice for the Australia quantum ecosystem, having been elected Australian of the Year in 2018. She showed the advances in her atomic-scale positioning technique for fabrication of spin qubits with single dopants in silicon.
Finally, the CEO of Cicada Innovations, Sally-Ann Williams, chaired two consecutive panels on the start up scene. The first one, formed by scientists and engineers, discussed the shift in view needed in order to translate tech-driven innovations to user-driven businesses. The second one saw venture capitalists from Main Sequence, Allectus/ICM and OpenOcean and the Executive Director of Tech Central (a Sydney precinct uniting several tech-oriented institutions, including the Quantum Terminal). They discussed the appetite that investors currently have for quantum innovation, the challenges associated to funding deep tech and the growth of venture capital in the Australian quantum ecosystem. Their main concerns included the difficulty with IP arrangements with universities (unlike North American universities that have a lot more experience handling commercialisation of high-value IP) and the lack of experience in constructing an investor thesis and pinpointing the financial bottom line for investors (even if it is based solely in valuation, instead of revenue).
The last day had no in-person component and consisted of a Careers Fair. A panel opened the day discussing the challenges in creating a quantum workforce, including the integration of professionals with different trainings, the focused upskilling of engineers and computer scientists, the issue of diversity and the space for non-PhD experts on this field (including those stemming from programs such as the new UNSW Quantum Engineering bachelor). Then, all sponsors were given space to describe their workplaces and opportunities for quantum experts in their organisations. The day ended with a short poster session – which unfortunately meant the conference was strongly focused on the leaders in the field and did not give much space for the next generation to showcase their work and views.
Dr. Saraiva has worked for over a decade providing theoretical solutions to problems in silicon spin quantum computation, as well as other quantum technologies. He works on MOS Silicon Quantum Dot research and commercially-oriented projects at the University of New South Wales (UNSW).
February 26, 2022