A key focus area where SandboxAQ has been applying their AI+Quantum (AQ) technology for some time is an area we call computational chemistry. This involves simulating the quantum mechanical effects of molecular structures to understand how they will react saving the need to empirically test it in the lab. The company has previously announced partnerships with NOVONIX and the University of San Francisco’s Institute for Neurodegenerative Diseases to simulate battery chemistries and accelerate drug discovery respectively.
SandoxAQ is now announcing that they will be collaborating with NVIDIA to use NVIDIA’s GPUs to accelerate some of the tensor network based algorithms used in these simulations. Specifically, NVIDIA will be able to utilize up to 32 NVIDIA H100 Tensor Core GPUs along with NVIDIA’s cuTENSOR and cuQuantum software to achieve the high performance needed to provide useful results in a reasonable period of time. There is wide applicability and demand for being able to simulate materials efficiently. Some specific examples of problems where SandboxAQ indicates this can be applied include protein-ligand binding computations for large-scale undruggable targets in neurodegenerative disease, novel solutions for pose and toxicity prediction, and predicting lifecycles for next-generation batteries composed of novel materials. Along with using the NVIDIA hardware and software, SandboxAQ will also provide technical recommendations to NVIDIA libraries that can be used for useful for future versions of these products.
A press release provided by SandboxAQ announcing their collaboration with NVIDIA can be accessed here.
November 23, 2023