[Q-tangled 22] Quantum Computing Timelines - Peter Morgan
0 likes42 views
The document discusses the current state and future potential of quantum computing, highlighting technological capabilities and the challenges faced in developing fault-tolerant quantum computers. It emphasizes the need for significant advancements in qubit quantity, error rates, and processing speeds to achieve commercial viability, projected to take 10-20 years. The analysis includes a SWOT assessment and cautions against overly optimistic timelines and regulatory risks.
[Q-tangled 22] Quantum Computing Timelines - Peter Morgan
- 1. Quantum Computing Timelines © Peter Morgan DSC Conference Nov 2022 Peter Morgan CEO Deep Learning Partnership
- 2. Quantum Computing Why now? • Physical phenomena obey the laws of • Classical physics • Quantum physics • Ultimately, at the deepest level, the universe is quantum • Systems are described by their wave functions • Properties of superposition and entanglement • If we can harness laws of quantum physics, we can build computers more powerful than classical computers to solve certain problems • We are at point in history where we can control objects at the quantum level • Individual atoms, ions and photons • “Atoms are Nature’s qubits” © Peter Morgan DSC Conference Nov 2022
- 3. Entanglement • Spooky action at a distance – Albert Einstein • I would not call entanglement one but rather the characteristic trait of quantum mechanics, the one that enforces its entire departure from classical lines of thought – Ernest Schrodinger © Peter Morgan DSC Conference Nov 2022
- 4. Quantum Technologies Quantum Computing Quantum Communications Quantum Sensing © Peter Morgan DSC Conference Nov 2022 • All three areas have long established and mature classical counterparts • Quantum sensing products are available now • Computing and communications at the PoC stage • We will talk about quantum computing – first proposed ~1980
- 5. Quantum Computing Components • QPUs • Control Systems • Lasers, MW pulses, software • Processor Interconnects • Must preserve entanglement • Quantum OS/Compiler • Application Software © Peter Morgan DSC Conference Nov 2022
- 6. Quantum Computing Hardware (QPUs) © Peter Morgan DSC Conference Nov 2022 • Any system with quantum energy levels that can be controlled • Currently have ~100 noisy qubits (need ~1million) • Timescales ~10-20 years before real commercial value? • NISQ à FTQC (Fault Tolerant )
- 7. Applications & Market • Simulation (materials, pharma, chemistry, energy) • Optimization (all industries) • Machine Learning (all industries) • Encryption (communications) --------------------------------- • Market for each is $trillions • Hugely impactful technology (in principle) © Peter Morgan DSC Conference Nov 2022
- 8. Path to Fault Tolerant Quantum Computers © Peter Morgan DSC Conference Nov 2022
- 9. Problems to Overcome What do we need from our quantum computers? • Tens of thousands to millions of qubits • Low error rates in qubits • Long qubit decoherence time • Fast gate speeds • High gate fidelities (low error rates) • Broad qubit interconnectivity • Need processor interconnects • Accurate readout of results (measurement) * Every vendor has problems with at least one of these – winners will overcome hurdles © Peter Morgan DSC Conference Nov 2022
- 10. © Peter Morgan DSC Conference Nov 2022 Mind the Gap or … “Beware the Quantum Chasm” We need to go from ~100’s to millions of qubits
- 11. Some Hardware Companies Azure Quantum © Peter Morgan DSC Conference Nov 2022
- 12. Some More Hardware Companies © Peter Morgan DSC Conference Nov 2022
- 13. Some Software Companies © Peter Morgan DSC Conference Nov 2022
- 14. When to Get Involved • If quantum advantage is 10 years away, is it - Too early to get involved? - Too late? - Right time? - How to time our involvement - Which companies to get involved with and when? - Let’s take a look! © Peter Morgan DSC Conference Nov 2022
- 15. What To Look For • Participants need to fully understand the physical and engineering challenges • What problems are we interested in solving? • What is quantum speedup over classical computer for such problems? • What circuit depth is needed to run the algorithms / simulations? • When might this be possible? • What hardware is best suited for a particular problem? • Utilize quantum computer benchmarks © Peter Morgan DSC Conference Nov 2022
- 16. Benchmarks Quantitative performance measures of a quantum computer – may include the following*: • Number of qubits • Connectivity • Gate fidelity • Gate speed • Decoherence time • Scaling • Measurement error *Industry is in the process of developing benchmarks – let’s look at some examples © Peter Morgan DSC Conference Nov 2022
- 17. Benchmarks (IBM) https://arxiv.org/abs/2110.14108 © Peter Morgan DSC Conference Nov 2022
- 18. Benchmarks (IonQ) © Peter Morgan DSC Conference Nov 2022 https://ionq.com/posts/february-18-2022-comparing-quantum- computers-metrics-monroney
- 19. Timeline Challenges • Not all the details are worked out yet for: • Algorithmic complexity (theoretical) • Compilers • Hardware • Interconnects • Error correction • Decoherence • Quantum many body physics • How to benchmark • Standards development • Can expect further experimental and theoretical breakthroughs © Peter Morgan DSC Conference Nov 2022
- 20. Market Analysis (SWOT) © Peter Morgan DSC Conference Nov 2022 Strengths = Powerful new form of compute Weaknesses = FTQC may not ever happen (or long time horizon) Opportunities = Massive speedups for certain impactful use cases Threats = A lot of significant challenges still to overcome Regulations could tighten Theoretical Regulatory Engineering
- 21. Hype Cycle © Peter Morgan DSC Conference Nov 2022 We are here? (Quantum winter?)
- 22. Conclusions © Peter Morgan DSC Conference Nov 2022 • Be realistic on timelines • Long time horizons (already lengthening) • Like fusion (always 20 years away)? • Have seen some acquisitions • No consensus yet on which hardware is best • Scalability is crucial for success • Beware the quantum winter • Crossing the quantum chasm • Recall AI has had three winters so far • Details Matter • Watch out for regulatory risk
- 23. Any Questions? © Peter Morgan DSC Conference Nov 2022