The Week in Quantum Computing - July 1st 2024 - Quantum Machines and Israel, Diraq, Gates vs Annealers, Community vs Quokkas. Pasqal goes up to 1000.
Issue #192
The Week in Quantum Computing. Brought to you by Sergio Gago (@piratecto).
Quick Recap
A Taiwanese delegation visited Bozeman, Montana, to explore collaboration opportunities in the photonics and quantum industries. The global quantum sensors market is projected to exceed $638.6 million by 2024. Moody's and QuEra are collaborating to improve tropical cyclone forecasting using quantum algorithms. QpiAI secured $6.5 million in pre-Series A funding to develop scalable quantum computers and enhance its software platforms, with plans to demonstrate a 25-qubit quantum computer by late 2024. Australian startup Diraq raised an additional $7 million to develop a CMOS quantum dot-based QPU, achieving a record 99.9% control accuracy for a qubit. DARPA's Quantum Benchmarking program identified over 200 applications for quantum computing, narrowing them to 20 benchmarks, but significant challenges remain in fields like computational fluid dynamics. Pasqal successfully trapped over 1,000 atoms in a single shot, marking a step towards scalable quantum processors. IonQ showcased its Forte Enterprise quantum computer at "Quantum Korea 2024," highlighting collaborations with Hyundai Motors and Korean universities.
Shaky times in Quantum
Guys, give me a break. This week many folks in the US will be off, so it may be a great time to reflect and think how we drive this industry forward and collaborate together. Two big beefs happened. One between an anonymous (but well known) X account and the Quokka device company and Chris Ferrie. And another between D-Wave and Q-Ctrl based on a paper published last week. Lets dive in:
Is the Quokka a quantum device? (spoiler alert: Who cares)
I am the first one who enjoys snarky comments against unproven claims. You, my dear readers, know that I try to take a neutral-optimist stance (or chaotic neutral if you are more of a D&D nerd). But this does not help. Chris has done more than most to share the magic and beauty of quantum through the world (even if we only consider the Quantum for babies book). His latest book/paper, Quantum Bullshit, is a fantastic divulgative piece. And his new company Quokka, builds a device that simulates quantum systems and can be used to teach more people and bring them in this beautiful world.
This is a flywheel folks. Researchers and other academics push the boundaries of what humans know. Some of them see potential benefit for others and build companies. Business folks take it and try to educate the world about it. Sometimes, taking some creative licenses (superposition is being in two states at the same time…), but that in turn creates more interest, gives more funding to academia, more grants, more postdocs, etc. Remember, some years ago, if you wanted to make good money with a PhD in physics most likely you had to go into finance. Today (fortunately) that is not the case anymore. That is because you are building awesome stuff, and some people knows how to sell it.
I would just say one thing (and to my college Electronics II professor as well). You don’t need hate, you can work together. If anything, the industry that feeds you is getting bigger. We are all trying to get to the same place.
The D-Wave - Q-Ctrl benchmark
Some of you may remember last week’s paper from Q-Ctrl claiming gate-based advantage over annealers for spin based problems. Of course D-Wave has answered with a paper here:
https://arxiv.org/abs/2406.19351v1
The conversation is alive and kicking on Linkedin. You can follow here. But here’s an extract from Michael Biercuk, CEO of Q-Ctrl
We're glad D-Wave has taken interest in our work with a recent response on arXiv. First and most importantly, they actually confirm that our results can indeed outperform annealers. Unfortunately their response also contains a factor of ~100X discrepancy in calculating a newly introduced metric.
What's this all about? We recently showed that a gate-model quantum computer can close the performance gap with annealers for optimization.
D-Wave in a response (Arxiv: 2406.19351) rerun the problems we treated with new insights beyond the original authors, and add a new "TTS" (time-to-solution) metric that leads to a 100X overestimate for our results relative to the base hardware sample time.
D-Wave:
"To estimate Q-CTRL time per sample tsample, we refer to Ref. [1] Appendix A, which specifies 6000 samples taking 150 seconds of QPU time..."
From our manuscript:
"The smallest problems we solved (28–32 nodes MaxCut) implemented 12 optimization steps, which totals to 72 circuits, each executed with 6,000 shots to a total of 432,000 shots throughout the full optimization process. IBM’s reported repetition time is ∼ 4, 000 shots per second, resulting in an ideal QPU time of 108 seconds. In practice, the measured QPU time of these jobs is about 150 seconds."
Thus the calculation divides 150s/6000 when a more natural hardware sample time is 150s/432,000.
We do not comment on the validity of the TTS metric overall as it appears it may be defined in such a way as to asymmetrically renormalize the IBM sample time. We remind that made no claims about comparative runtime advantages in our work, only that gate-model machines could for the first time outperform annealers.
We congratulate the team on developing a new implementation of the problem which appears to perform better than previous pubs. This is well aligned with our expectations:
"The results presented in this work demonstrate that an appropriately constructed quantum solver executed on an IBM gate-model quantum computer can outperform published outputs from quantum annealers for nontrivial binary optimization problems. Such a statement warrants careful qualification...as we relied upon published problem instances and implementations for quantum annealers we do not exclude the possibility of crafting better solutions in detailed implementation... [previous publication Ref 19] contains no optimality proof and an alternate implementation could deliver superior annealer results. Further, our results do not preclude the existence of alternate problem statements in which annealers retain advantages over gate-model machines, or that individual hardware systems may undergo periodic upgrades to deliver improved performance relative to the reference results used here. Overall, despite these qualifications we believe these results now challenge the heretofore practically correct assertion that annealers would consistently outperform gate-model machines for quantum optimization problems."
I would like to take a moment to reflect on those two situations. One of them has a scientific based debate (heated, but very interesting) both in the shape of papers and Linkedin. Some may say LinkedIn is not the right place for scientific debate… whatever… <sarcasm>The only way to produce relevant and scientifically accurate content is by publishing a non peer-reviewed paper on a website</sarcasm>
Anyway, two interesting situations, dealt in different ways, that create even more friction in the industry.
Now, for what actually happened during the week (very relevant stuff!)
The Week in Quantum Computing
DARPA Releases Preliminary Findings of Quantum Computer Benchmarking Research Effort
DARPA’s Quantum Benchmarking program has published preliminary results of its scientific efforts to measure the impact of quantum computers on key computational challenges. The program's phase 2 involved teams from University of Southern California, HRL Laboratories, L3Harris Technologies, Rigetti Computing, and Zapata Computing. The study focused on three categories: chemistry, materials science, and non-linear differential equations. Initial results suggest potential advantages of using quantum computers for chemistry and materials science applications, while its benefits for non-linear differential equations remain unclear. Quantum Benchmarking Program Manager, Joe Altepeter, noted these findings as an important first step towards quantifying the impact of quantum computers.
Taiwan, Montana seek to build cooperation in photonics, quantum industries
Taiwan and Montana are seeking to build cooperation in the photonics and quantum industries. A delegation of representatives from Taiwan's photonics and quantum industries recently visited Montana, meeting with Governor Greg Gianforte and Montana Department of Commerce Director Paul Green. They also visited technology research centers and companies such as the Tech Hub at Montana State University, Vision Aerial, and CACI. Montana has one of the highest per capita concentrations of optics, photonics, and quantum computing companies in the U.S., while Taiwan has a significant optoelectronics industry. The visit aimed to boost cooperation in these sectors.
https://focustaiwan.tw/business/202406230008
Quantum Sensors Market Set to Surpass $638.6M by 2024, Pioneered by MIT, IBM, and Google Amidst Skill Shortage Challenge
The global quantum sensors market is projected to exceed US$638.6 million in 2024, driven by advancements in quantum technology and research, according to Research and Markets. Key players such as MIT, IBM, and Google are investing heavily in quantum research, leading to the development of highly sensitive magnetometers, gravimeters, and atomic clocks. These innovations are crucial for various applications, including medical imaging and navigation. However, the market faces challenges due to a shortage of professionals skilled in quantum computing, nanotechnology, and advanced physics.
https://www.communicationstoday.co.in/quantum-sensors-market-to-surpass-us638-6-million-in-2024/
TQI Exclusive: Moody’s, QuEra Explore Using Quantum Computers to Better Predict Path, Intensity of Tropical Cyclones
Moody’s and QuEra have partnered to explore the use of quantum algorithms in predicting the path and intensity of tropical cyclones. The team aims to create a quantum reservoir computing-based model to process climate model outputs and storm environment parameters for more accurate forecasting. This could improve economic impact predictions and disaster preparedness. Ricardo Garcia, a quantum computer engineer at Moody’s Analytics, highlighted the commercial interest in assessing the risk of extreme events like hurricanes. The National Oceanic and Atmospheric Administration estimated that Hurricane Katrina caused over $125 billion in damage.
QpiAI Closes $6.5 Mill Pre-Series A Funding Led by Yournest and SIDBI Venture Capital to Enable Intelligence Modelling and Intelligence Compute Using Quantum Computers
QpiAI, a leader in quantum computing and generative AI, has secured $6.5 million in pre-Series A funding led by Yournest and SIDBI Venture Capital. Other participants included WFC, an angel group, and numerous individual investors. The investment will be used to develop full stack 25 qubit Quantum computers, scalable up to 1000 physical superconducting qubits. QpiAI aims to increase revenues from software licensing of seven platforms based on Quantum computing and Generative AI technologies. The company anticipates demonstrating 25 Qubit Quantum computers by Q4 CY2024/ Q1 CY 2025.
https://www.businesswirechina.com/en/news/57097.html
DARPA searched for fields quantum computers really could revolutionize, with mixed results
The US Defense Advanced Research Projects Agency (DARPA) has published results of its Quantum Benchmarking program, which aimed to measure quantum computing progress and its possible future applications. The program included eight interdisciplinary teams which identified over 200 potential applications. Detailed studies were conducted in three broad categories: chemistry, materials science, and non-linear differential equations. The results, published in seven pre-press papers, suggest that quantum computers could provide advantages for applications in certain chemistry, quantum materials, and materials science areas. However, they also highlighted challenges, particularly in the field of incompressible computational fluid dynamics (CFD) simulations. The papers suggest that quantum computers could bring about $2 million savings for each test at the Los Alamos National Laboratory's High Magnetic Field Laboratory.
https://www.theregister.com/AMP/2024/06/24/darpa_quantum_computer_benchmarking_papers/
Diraq boosts CMOS quantum dot funding round to $22m
Australian quantum tech startup Diraq has raised its latest investment round by $7m to $22m, aiming to develop a quantum processing unit (QPU) using CMOS quantum dot technology for a fault-tolerant quantum computer with millions of qubits. The funding round included strategic investors Main Sequence, Taronga Ventures, Uniseed, UniSuper, Co:Act Capital, and foundation shareholder and research partner UNSW, Sydney, and was led by Quantonation. Diraq recently announced a control accuracy of 99.9% for a qubit manufactured by imec using industry-standard CMOS materials, demonstrating the potential scalability of the technology.
https://www.eenewseurope.com/en/diraq-boosts-cmos-quantum-dot-funding-round-to-22m/
Pasqal Achieves Key Milestone Exceeding 1,000 Atoms in Quantum Processor, Paving Way for Scalable Quantum Computing
Quantum computing leader, Pasqal, has successfully loaded over 1,000 atoms in a single shot within their quantum computing setup, marking a significant technological milestone. The company trapped more than 1,110 atoms within approximately 2,000 traps, demonstrating the potential for large-scale neutral atom quantum computing. The achievement involves innovative optical designs that combine ultra-high-vacuum-compatible microscope objectives with windowless thermal shields for efficient trapping at cryogenic temperatures. Co-CEO of Pasqal, Loic Henriet, stated that these results will fuel the design of future hardware products with enhanced computational power. This milestone aligns with Pasqal's strategic roadmap, which targets the development of quantum computers with over 1,000 qubits, progressing towards 10,000 qubits by 2026-2027.
IonQ Showcases Strong Partnerships at Quantum Korea 2024
IonQ showcased its partnerships and latest quantum computer, Forte Enterprise, at the Quantum Korea 2024 event. The company highlighted collaborations with Hyundai Motors, Seoul National University, Sungkyunkwan University, and others. The Forte Enterprise, with a target performance of 36 qubits, is designed for complex computing tasks. Hyundai Motors uses IonQ's quantum computers for battery development and autonomous driving research. IonQ has renewed its partnership with Sungkyunkwan University and signed an agreement with Seoul National University to foster quantum talent.
https://ionq.com/news/ionq-showcases-strong-partnerships-at-quantum-korea-2024
Quokka: World's first personal quantum computer emulator that fits in a hand
Researchers from the Centre for Quantum Software and Information (QSI) have developed Quokka, the world's first personal quantum computer emulator. Co-founded by Simon Devitt and Chris Ferrie, Quokka aims to make quantum computing accessible and understandable to everyone. The device, which is set to ship in July, emulates a 30-qubit fault-tolerant quantum computer and can run quantum computing programming languages. The founders believe that quantum literacy will define 21st-century innovation and hope to revolutionize STEM education. Ferrie states that Quokka allows users to experiment and learn about quantum algorithms and programs, providing a hands-on experience with cutting-edge technology.
https://interestingengineering.com/innovation/quokka-personal-quantum-computer
Orca to lead R&D consortium developing multiplexing technologies for quantum computers
Quantum computing company Orca has announced it will lead an R&D consortium to develop network multiplexing and switching technologies for use in quantum computing and data centers. The consortium also includes Toshiba Europe, Bay Photonics, Imperial College London, and University College London. Multiplexing and switching are foundational technologies used to enable ultra-high data transmission rates and are a necessary puzzle piece in the scaling up of quantum systems beyond the limits of individual quantum processors.
Quantinuum researchers tackle AI’s interpretability problem, helping us build safer systems
Quantinuum researchers are addressing the 'interpretability problem' in AI systems, which currently makes them difficult to understand or interpret. They have recently published a paper outlining the issue. The lack of explainability in machine learning models comes from their underlying architecture. This is a major concern in areas where accountability is needed, such as finance and healthcare. Researchers are working on new machine learning models that are inherently interpretable from the start. The paper suggests that such models could eliminate the need for explainable AI methods. Quantinuum's recent paper presents a new theoretical framework for defining AI models and analyzing their interpretability, focusing on the advantages of compositional structure. The research is part of Quantinuum's broader AI strategy, which includes using AI to improve quantum computing and vice versa.
Paper: Towards Compositional Interpretability for XAI
EY and Oxford University Study Pushes for Responsible and Ethical Approach to Quantum Computing
A study by EY and Oxford University's Responsible Technology Institute (RTI) calls for a responsible approach to quantum computing as it transitions towards commercial reality. The research emphasizes the balance between innovation and responsibility, urging accurate and transparent communication to avoid overhyping the technology and causing public distrust. Digital Ethics Lead at EY, Mira Pijselman, stresses the importance of counteracting misleading information. The study also underscores the necessity of involving diverse groups in the technology's development, with 87% of respondents supporting this view. Furthermore, 92% support government involvement in funding new technology.
https://technologymagazine.com/articles/ey-the-call-for-responsible-innovation-in-quantum-computing
Evaluating cat qubits for fault-tolerant quantum computing using Azure Quantum Resource Estimator
Microsoft and French startup Alice & Bob have teamed up to evaluate cat qubits for fault-tolerant quantum computing using Microsoft's Azure Quantum Resource Estimator. The tool, made open-source in January 2024, can now model any quantum architecture and modify all assumptions. Alice & Bob's cat qubit architecture requires a unique error correction strategy due to its biased noise, necessitating the modification of the usual gate hierarchy in quantum computing. The collaboration demonstrated the estimator's extensibility by modeling Alice & Bob's architecture, demonstrating its utility in assessing the practical benefit of quantum algorithms.
Q-Day countdown: HSBC Unveils Strategies to Secure Banking Systems
HSBC's Global Head of Quantum Technologies, Philip Intallura, has warned of the cybersecurity risks posed by quantum computing, particularly to the financial sector. With HSBC processing around 4.5 billion transactions annually for 41 million customers, the potential for quantum computers to break public key cryptography is a serious concern. The bank is developing internal initiatives and partnerships to safeguard customer data against the anticipated 'Q-Day' when quantum computers can break existing cryptographic algorithms. HSBC is exploring post-quantum cryptography (PQC) and quantum key distribution (QKD) as protective measures. The bank has partnered with BT and Toshiba on a quantum-secured metro network, and has simulated a 30m euro to dollar foreign exchange trade on the network.
https://www.infosecurity-magazine.com/interviews/q-day-hsbc-secure-banking/
South Korean Government, NH Bank and OurCrowd to Establish $80 Million Israel-Korea Fund for DeepTech Investments
The South Korean government, in collaboration with NH Bank and Israel's leading venture investor, OurCrowd, is establishing an $80 million Israel-Korea fund. This bilateral investment fund is designed to support startups in both countries, with a particular focus on DeepTech and quantum investments. OurCrowd's selection by the South Korean government to co-manage this fund highlights its prominence in the venture investing landscape.
Quantum Computing Takes Off With $55 Billion In Global Investments
Global investments in quantum computing have reached $55 billion, bringing the world closer to realizing the $500 million to $1 billion in gains the technology promises to deliver to businesses over the next 15 years. Notable investments include Germany's $3 billion plan by 2026 and France's nearly $2 billion investment aimed at training 5,000 quantum-ready engineers and creating 30,000 jobs. Quantum computers, potentially a billion times faster than conventional computers, use quantum bits, or qubits, capable of existing in multiple states simultaneously. IBM, a leader in quantum computing hardware, predicts that quantum computers will outperform classical computers in specific tasks by 2027. The quantum market is already estimated to be worth more than $1 billion this year.
https://www.forbes.com/sites/sylvainduranton/2024/06/26/quantum-now/
Paper: A comment on comparing optimization on D-Wave and IBM quantum processors
A recent study by McGeoch, Chern, Farré, and King challenges Q-CTRL's claim of superior performance of IBM's quantum processing units (QPUs) over D-Wave's quantum annealer. The authors argue that Q-CTRL's comparison methodology was flawed and demonstrate that a simple, unoptimized workflow for quantum annealing yielded success probabilities multiple orders of magnitude higher than those reported by Q-CTRL. Furthermore, their direct comparison between quantum annealing and digitized quantum annealing showed that D-Wave's QPU achieved far lower energies than IBM's QPU.
https://arxiv.org/abs/2406.19351v1
Call for new DOD-led quantum hub builds momentum on the Hill
Rep. Nancy Mace is proposing a $20 million increase in Army RDT&E funding to establish a first-of-its-kind Quantum Computing Center of Excellence, aimed at aiding the military. This amendment to the Defense Department’s appropriations bill for fiscal 2025 is part of a broader effort by the U.S. national security agencies to prioritize quantum-enabling activities and funding. The investment will be offset by a decrease in Defense-Wide Operations and Maintenance. The House Armed Services Subcommittee on Cyber Innovative Technologies and Information Systems (CITI) supports the establishment of this quantum center.
https://defensescoop.com/2024/06/21/call-for-new-dod-led-quantum-hub-builds-momentum-on-hill/