Gregory. D. Peterson

Artificial intelligence computational capabilities are emerging as an important capability for research across many disciplines and having access to these capabilities is also a key component in research training for developing the future, skilled AI workforce. The University of Tennessee, Knoxville started the AI Tennessee Initiative in September 2022 specifically to support AI related research and research training. The UTK AI Tennessee Initiative provided funds to purchase seven Dell PowerEdge XE8640 high performance computing servers to support the Initiative. The XE8640 server was new from Dell for 2023 and is an air cooled HPC class system capable of supporting up to four high end GPUs. The servers were placed into UTK’s ISAAC Next Generation campus cluster. This paper describes the ISAAC NG cluster and infrastructure, acquisition and installation of the XE8640 hardware, benchmarking, power consumption, and analysis of the benchmark results, as well as lessons learned that may be applicable to other systems.

Biomedical machine reading comprehension (bio-MRC), a crucial task in natural language processing, is a vital application of a computer-assisted clinical decision support system. It can help clinicians extract critical information effortlessly for clinical decision-making by comprehending and answering questions from biomedical text data. While recent advances in bio-MRC consider text data from resources such as clinical notes and scholarly articles, the clinical practice guidelines (CPGs) are still unexplored in this regard. CPGs are a pivotal component of clinical decision-making at the point of care as they provide recommendations for patient care based on the most up-to-date information available. Although CPGs are inherently terse compared to a multitude of articles, often, clinicians find them lengthy and complicated to use. In this paper, we define a new problem domain – bio-MRC on CPGs – where the ultimate goal is to assist clinicians in efficiently interpreting the clinical practice guidelines using MRC systems. To that end, we develop a manually annotated and subject-matter expert-validated benchmark dataset for the bio-MRC task on CPGs – cpgQA . This dataset aims to evaluate intelligent systems performing MRC tasks on CPGs. Hence, we employ the state-of-the-art MRC models to present a case study illustrating an extensive evaluation of the proposed dataset. We address the problem of lack of training data in this newly defined domain by applying transfer learning. The results show that while the current state-of-the-art models perform well with 78% exact match scores on the dataset, there is still room for improvement, warranting further research on this problem domain. We release the dataset at https://github.com/mmahbub/cpgQA.

Mixed precision is a promising approach to save energy in iterative refinement algorithms since it obtains speedup without necessitating additional cores and parallelisation. However, conventional mixed precision methods utilise statically defined precision in a loop, thus hindering further speed-up and energy savings. We overcome this problem by proposing novel methods which allow iterative refinement to utilise variable precision arithmetic dynamically in a loop (i.e. a trans-precision approach). Our methods restructure a numeric algorithm dynamically according to runtime numeric behaviour and remove unnecessary accuracy checks. We implemented our methods by extending one conventional mixed precision iterative refinement algorithm on an Intel Xeon E5-2650 2GHz core with MKL 2017 and XBLAS 1.0. Our dynamic precision approach demonstrates 2.0–2.6× speed-up and 1.8–2.4× energy savings compared to mixed precision iterative refinement when double precision solution accuracy is required for forward error and with matrix dimensions ranging from 4K to 32K.

The Computational Science for Undergraduate Research Experiences (CSURE) is an NSF funded Research Experiences for Undergraduates (REU) program organized by the University of Tennessee's Joint Institute for Computational Sciences (JICS), www.jics.utk.edu/csure-reu. The main goal of the CSURE project is to direct a group of undergraduate students to explore the emergent computational science models and techniques using the supercomputers at the National Institute for Computational Sciences (NICS). In addition, a number of summer interns from Hong Kong also participate in the program. The CSURE program focuses on five different scientific domains: chemistry and material sciences, systems biology, engineering mechanics, atmospheric sciences, and parallel solvers on emergent platforms. The program also enjoys a close affiliation with researchers at the Oak Ridge National Laboratory. Because of these diverse topics of research areas and backgrounds of this project, in this paper we discuss the challenges and resolutions in managing and coordinating the program, delivering cohesive tutorial materials, directing mentorship of individual projects, and lessons learned in the duration of the program.

Computing in science and engineering is now ubiquitous: digital technologies underpin, accelerate, and enable new, even transformational, research in all domains. Access to an array of integrated and well-supported high-end digital services is critical for the advancement of knowledge. Driven by community needs, the Extreme Science and Engineering Discovery Environment (XSEDE) project substantially enhances the productivity of a growing community of scholars, researchers, and engineers (collectively referred to as "scientists"' throughout this article) through access to advanced digital services that support open research. XSEDE's integrated, comprehensive suite of advanced digital services federates with other high-end facilities and with campus-based resources, serving as the foundation for a national e-science infrastructure ecosystem. XSEDE's e-science infrastructure has tremendous potential for enabling new advancements in research and education. XSEDE's vision is a world of digitally enabled scholars, researchers, and engineers participating in multidisciplinary collaborations to tackle society's grand challenges.

Background: The currently known protein sequence space consists of millions of sequences in public databases and is rapidly expanding. Assigning sequences to families leads to a better understanding of protein function and the nature of the protein universe. However, a large portion of the current protein space remains unassigned and is referred to as its “dark matter”. Results: Here we suggest that true size of “dark matter” is much larger than stated by current definitions. We propose an approach to reducing the size of “dark matter” by identifying and subtracting regions in protein sequences that are not likely to contain any domain. Conclusions: Recent improvements in computational domain modeling result in a decrease, albeit slowly, in the relative size of “dark matter”; however, its absolute size increases substantially with the growth of sequence data.