Happy to see this one out! Ever wonder what we can do with a Rydberg atom array Quantum Processing Unit? Check out our new preprint: "Leveraging Analog with Neutral Atoms for Solvent Configuration Prediction in Drug Discovery". A collective effort involving Qubit Pharmaceuticals, Pasqal and our group. arxiv.org/abs/2309.12129

Please RT.
2 PhD positions are available in my group in the context of the PhyMol EU project.
- DC5 in collaboration with Alexandre Tkatchenko (UL, Luxembourg);
- DC6 with Alston Misquitta (QMUL, UK).
=> All application details are on the website. phymol.eu/research.php

Our recent paper: "Overlap-ADAPT-VQE: practical quantum chemistry on quantum computers via overlap-guided compact Ansätze" made it to the front page of Communications Physics (Nature porftolio). Check it out:

twitter.com/CommsPhys/status/1

🔥Hot off the press🔥
Our paper introducing the Overlap-ADAPT-VQE method to grow wave-functions by maximizing their overlap with any intermediate target wave-function is now online in Communications Physics. Sorry for the barren plateaus!!!😃
Shout out to the amazing César Feniou & Muhammad Hassan. Great interdisciplinary collaboration with Yvon Maday.
rdcu.be/dh2QY

Huge work by César Feniou and great collaboration with Yvon Maday's group (LJLL@Sorbonne). Funded in part @ERC_Research (project EMC2). A big thank to Amazon Braket and to Qubit Pharmaceuticals for support.

Show thread

Preprint:Lambda-ABF: Simplified, Accurate & Cost-effective Alchemical Free Energy Computations. New Colvars library for NAMD & Tinker-HP. Great efforts by L. Lagardère & L. Maurin and collaborations withs P. Monmarché & J. Hénin
arxiv.org/abs/2307.08006

4/4 · We also show a 1st implementation of a resource-efficient Overlap-ADAPT-like approach (arXiv:2301.10196). The hybrid approach is shown to provide high fidelities. 2 approaches for the quantum estimation of overlaps are provided: SWAP-Test & the Compute-Uncompute method.

Show thread

3/4 · We show a successful simulation of a 25-spin Ising model through an adaptive hybrid quantum/HPC method exhibiting a quantum complexity independent of the number of qubits & extensible to other physical model Hamiltonians. The approach is shown to provide high fidelities;

Show thread

2/4 we introduce a new class of “hardware-friendly” adaptive methods: the Greedy Gradient-free Adaptive-VQE (GGA-VQE) and provides the 1st implementation of an Adaptive-Variational Quantum Eigensolver (Adaptive-VQE) on a 25-qubit error mitigated quantum computer (IonQ Aria)

Show thread

New : Smooth Particle Mesh Ewald-integrated stochastic Lanczos Many-body Dispersion algorithm. Great work by Pier Poier. Funding ERC_Research (project EMC2) Supercomputer time by GENCI.
arxiv.org/abs/2307.02278

Stellar work by Igor Chollet. Funding @ERC_Research (project EMC2). @LCT_UMR7616

Show thread

Just out in JCTC : ANKH: A Generalized O(N) Interpolated Ewald Strategy for Molecular Dynamics Simulations. ANKH remains scalable for any size of systems toward exascale & generalized for distributed point multipoles & induced dipoles doi.org/10.1021/acs.jctc.3c000

Impressive work by Igor Chollet. Funding @ERC_Research (project EMC2). Supercomputing time @Genci_fr. @LCT_UMR7616

Show thread

Just out in JCTC @JCIM_JCTC: ANKH: A Generalized O(N) Interpolated Ewald Strategy for Molecular Dynamics Simulations. ANKH remains scalable for any size of systems toward & generalized for distributed point multipoles & induced dipoles. doi.org/10.1021/acs.jctc.3c000

RT @HPC_Guru
With BF16, @nvidia H100s are both faster and more cost effective than A100s for training

With FP8, H100s were 30% more cost-effective and 3x faster than A100s for a 7 billion parameter MosaicGPT model

via @MosaicML @ProfMatsuoka twitter.com/MosaicML/status/16

Show older
Mastodon Sorbonne Université

Serveur Mastodon de Sorbonne Université