Quantum Thursdays Q&A – September Edition
Our expert speakers answer your questions from the September edition of Quantum.Tech's Quantum Thursdays. Take a look at their answers here:
What are the uses of Solving Differential Equations for the speedup of super polynomials as proposed for quantum computing (#Q developer), and within financial services?
Solving differential equations has lots of applications in hard engineering and scientific problems. In financial problems, it also has applications in e.g. option pricing (Black-Sholes model and alike).
Which one is most in use in pharmaceuticals? Variation Quantum Eigensolver or Trotter Suzuki for solution to a quantum system?
Typically VQE, because it’s relatively easy to program on current quantum processors and it allows you to search for optimal molecule configurations. The Trotter-Suzuki quantum simulation would be ideal for e.g. time evolution, but it’s typically harder to code and may involve more quantum gates.
Do you have roadmap to test the quantum algorithms from DWave annealers in Q4 2020 with 5000% 2B qubits to Fujitsu new annealer in Q4 2020? In 2021 with 128 Qubits by IBM and Honeywell 640 Qubits?
Our roadmap is to test these processors as soon as they become available.
Why strive to construct Hamiltonians for finance problems and not derive them from Lagrangians? This way least-action principle can be implemented when capital is moving from one investment to another in analogy to energy converted from potential to kinetic. Minimization (or maximization) this way is natural.
This is true: the portfolio optimization problem can be seen in its continuous version as a minimum-action problem, leading to some Lagrange-euler equations. The problem is then mapped to a system of coupled differential equations, which is also hard to solve in some cases. This could also be tackled via quantum and quantum-inspired methods.
Will France get their own quantum computers like in Germany, IBM and Frauenhofer to enable the industry in quantum computing local?
French companies are developing interesting hardware, for example Pasqal (cold atoms) or plan to develop groundbreaking hardware (Alice & Bob, aiming for the first fault tolerant QC). It’s unclear whether there will be a need to buy and operate hardware like a Germany university is doing with IBM – cloud access is still the preferred method.
On quantum distribution input, have you seen last JPMC paper?
It’s interesting using heuristical method like VQE to do so. It’s a good step forward, but still doesn’t guarantee the initial state prepared that way is exactly what was designed. I hope we’ll have hardware methods to do that with accuracy in the future.
What happens when the AI Chip cerebras is combined with a quantum computer?
It could happen between 2025-2035. It’s a high range of possibilities, but we see improvements on the “classical” hardware side with TPU and neuromorphic computing that could delay quantum advantage in AI.
Catch up on our monthly digital series here.