「NVIDIA GTC 2023 (GPU Technology Conference)」にて講演予定

Preferred Networks エンジニアのZijian Xu が、2023年3月21日（火）~ 24日（金）に開催される「NVIDIA GTC 2023 (GPU Technology Conference)」にて講演いたします。

10万原子という大きな構造をNeural Network Potentialで扱うために、Recomputation (再計算)と呼ばれる手法を用いてメモリ使用量を効率化する、という内容です。

詳細はこちらをご覧ください。無料で参加登録をしていただくことで、講演・録画を見ることができます。
https://www.nvidia.com/gtc/session-catalog/?tab.catalogallsessionstab=16566177511100015Kus&search=running#/session/1666600863417001yG53

講演の詳細

日時	2023年3月22日（水）　12:00-12:25 JST
講演者	Preferred Networks エンジニア　Zijian Xu
場所	オンライン
タイトル	Running 100,000-Atom Molecular Dynamics with Accurate NN Potential on a GPU using Automatic Recomputation
講演概要	Accurate and large scale atomistic simulation allows a new way of material discovery. Although neural network potential (NNP), a deep learning-based force field, dramatically accelerates computation time compared to conventional DFT methods, the amount of GPU memory often becomes a bottleneck to running large-scale simulations. Supporting a larger number of atoms in such simulations greatly increases the range of structures and phenomena that can be reproduced, including large molecules like proteins, heterogeneous systems like composites, additives, and lattice-mismatched interfaces. To overcome this issue, our method analyzes and reorganizes the computation graph of the neural network model to achieve exactly the same outputs with a smaller memory budget. The basic idea here is known as “recomputation,” which wipes out less frequently referenced large data from GPU memory and recomputes them from smaller values when necessary. Our experiments have shown that it is possible to run up to 5x larger simulations with only a 25% execution time overhead using the same amount of GPU memory.