Engineering topics

Davis, James

Publications and source records attributed to Davis, James.

3 records

DOE OSTI2023

GenSLMs: Genome-scale language models reveal SARS-CoV-2 evolutionary dynamics

We seek to transform how new and emergent variants of pandemic-causing viruses, specifically SARS-CoV-2, are identified and classified. By adapting large language models (LLMs) for genomic data, we build genome-scale language models (GenSLMs) which can learn the evolutionary landscape of SARS-CoV-2 genomes. By pre-training on over 110 million prokaryotic gene sequences and fine-tuning a SARS-CoV-2-specific model on 1.5 million genomes, we show that GenSLMs can accurately and rapidly identify variants of concern. Thus, to our knowledge, GenSLMs represents one of the first whole-genome scale foundation models which can generalize to other prediction tasks. We demonstrate scaling of GenSLMs on GPU-based supercomputers and AI-hardware accelerators utilizing 1.63 Zettaflops in training runs with a sustained performance of 121 PFLOPS in mixed precision and peak of 850 PFLOPS. We present initial scientific insights from examining GenSLMs in tracking evolutionary dynamics of SARS-CoV-2, paving the path to realizing this on large biological data.

Zvyagin, Maxim↗

DOE OSTI2020

Cyberinfrastructure for the democratization of smart manufacturing

“Smart Manufacturing” is the transformation of U.S. manufacturing that results from disrupting traditional business, organizational, operating, and market structures with a radical increase in the availability and use of real-time operations data to produce value in previously inconceivable ways—a transformation process known as “manufacturing digitalization” or “manufacturing digital transformation.” There are significant economic and investment opportunities with substantially increased supply chain productivity and far better product design with process and machine precision and performance for highervalue products manufactured better, faster, cheaper with less energy and material.

Advanced sensor controls platforms and modeling↗

NASA NTRS2019

Evaluation of Ionic Liquids for Closed Air Revitalization

This document is the final report resulting from the work conducted by undergraduate students at the University of South Alabama during the 2018/2019 academic year and was prepared by the undergraduate students. As NASA pushes the boundaries further into space, the current technologies within the various life support systems must be improved upon. One such improvement is needed to the current air revitalization systems, specifically sorbents that can capture CO2 more effectively from enclosed habitats. Ionic liquids (ILs) have been considered as absorbents for flue gas, but little research has been done to test the ability of ILs at ambient pressures and relatively low concentration of CO2. The experiment outlined below utilizes the task-specific ionic liquid, tetramethylammonium taurinate (TMN), in a commercial off the shelf absorption system to capture CO2. The CO2 stream is combined with nitrogen to produce an inlet gas concentration relevant to close air revitalization applications. At an inlet gas flow with a CO2 partial pressure of 3.8 torr the system was capable of removing just under 97% of the inlet CO2. The concentration of CO2 in the outlet stream, partial pressure 0.16 torr, was less than that of atmospheric air. The duty required to separate the absorbed gas from the ionic liquid as well to cool the ionic liquid to be reintroduced to the column were acquired utilizing laboratory cooling/heating baths. These results show that TMN may be an efficient candidate for consideration in closed air revitalization.

Duck, Gerald↗