.svg){kind=small}
October 4, 2024
New funding will help scientists develop drugs for muscle wasting disease
https://www.htworld.co.uk/news/research-news/new-funding-will-help-scientists-develop-drugs-for-muscle-wasting-disease-dd24/
No items found.
Here are some scientific papers that caught our eye this month:
This paper discusses exponential speedup involving simulating coupled harmonic oscillations. Such simulation is fundamental to simulating molecular dynamics and has significant implications for materials science, chemistry, and drug discovery.
The researchers' approach for preparing the initial state requires exponentially-large resources. Despite this, the formal proof that a quantum computer can efficiently solve this problem (making it "BQP-complete") and its potential applications make this research noteworthy.
April was a big month for experimental results as well! In this collaboration, QuEra’s cofounders, and teams at Harvard, MIT, and QuEra, pushed the boundaries of state-of-the-art 2-qubit gate fidelities with neutral atoms. Entangling gates with 99.5% fidelity on up to 60 atoms in parallel were demonstrated, surpassing the threshold for surface code error correction. The results were also extended to demonstrate low-error three-qubit gates. This result is a landmark paper towards universal fault-tolerant quantum computers with neutral atoms.
Quantum error correction is crucial for creating practical, universal quantum computers. This paper discusses how we can connect current, less powerful quantum devices (NISQ devices) with future, error-resistant ones. The authors propose a new approach that avoids complex protocols and uses careful state management. Their method shows promise for achieving error resistance beyond what NISQ devices can handle, potentially with as few as 10,000 qubits. However, this approach also requires very low error rates in the qubits. The community is still evaluating whether this proposal is strong, but the idea of avoiding complex protocols is gaining interest, and this paper adds important details and estimates to that discussion.
Quantum machine learning is a popular field, but training quantum neural networks can be difficult. Reservoir computing is an alternative approach for quantum learning tasks. This work shows that quantum coherence can be useful in reservoir computing, even beating classical methods in some cases. Since training classical machine learning methods can be resource-intensive, quantum reservoir computing could be a valuable alternative and a significant application of quantum computing.
