A5012850684- Distributed and Parallel Computing Systems
- Parallel Computing and Optimization Techniques
- Advanced Data Storage Technologies
- Cloud Computing and Resource Management
- Scientific Computing and Data Management
- Computational Drug Discovery Methods
- Distributed systems and fault tolerance
- Software System Performance and Reliability
- Embedded Systems Design Techniques
- Software-Defined Networks and 5G
- Service-Oriented Architecture and Web Services
- Protein Structure and Dynamics
- Systems Engineering Methodologies and Applications
- Genetics, Bioinformatics, and Biomedical Research
- SARS-CoV-2 detection and testing
- Brain Tumor Detection and Classification
- Peer-to-Peer Network Technologies
- SARS-CoV-2 and COVID-19 Research
- Caching and Content Delivery
- Semantic Web and Ontologies
- Machine Learning in Materials Science
- Simulation Techniques and Applications
- Particle Detector Development and Performance
- Data Mining Algorithms and Applications
- Security and Verification in Computing
Oak Ridge National Laboratory
2012-2023
Center for Clinical Studies
2021
Summary The Exascale Computing Project (ECP) is currently the primary effort in United States focused on developing “exascale” levels of computing capabilities, including hardware, software, and applications. In order to obtain a more thorough understanding how software projects under ECP are using, planning use Message Passing Interface (MPI), help guide work our own project within ECP, we created survey. Of 97 active at time survey was distributed, received 77 responses, 56 which reported...
Time-to-solution for structure-based screening of massive chemical databases COVID-19 drug discovery has been decreased by an order magnitude, and a virtual laboratory deployed at scale on up to 27,612 GPUs the Summit supercomputer, allowing average molecular docking 19,028 compounds per second. Over one billion were docked two SARS-CoV-2 protein structures with full optimization ligand position 20 poses docking, each in under 24 hours. GPU acceleration high-throughput optimizations program...
Abstract This dataset contains ligand conformations and docking scores for 1.4 billion molecules docked against 6 structural targets from SARS-CoV2, representing 5 unique proteins: MPro, NSP15, PLPro, RDRP, the Spike protein. Docking was carried out using AutoDock-GPU platform on Summit supercomputer Google Cloud. The procedure employed Solis Wets search method to generate 20 independent binding poses per compound. Each compound geometry scored AutoDock free energy estimate, rescored RFScore...
We introduce UnifyFS, a user-level file system that aggregates node-local storage tiers available on high performance computing (HPC) systems and makes them to HPC applications under unified namespace. UnifyFS employs transparent I/O interception, so it does not require changes application code is compatible with commonly used libraries. The design of supports the predominant workloads optimized for bulk-synchronous patterns. Furthermore, provides customizable semantics flexibly adapt its...
The urgent search for drugs to combat SARS-CoV-2 has included the use of supercomputers. general-purpose graphical processing units (GPUs), massive parallelism, and new software high-performance computing (HPC) allowed researchers vast chemical space potential faster than ever before. We developed a drug discovery pipeline using Summit supercomputer at Oak Ridge National Laboratory help pioneer this effort, with platforms that incorporate GPU-accelerated simulation allow virtual screening...
We present a supercomputer-driven pipeline for
The SPEChpc 2021 suites are application-based benchmarks de- signed to measure performance of modern HPC systems. bench- marks support MPI, MPI+OpenMP, MPI+OpenMP target offload, MPI+OpenACC and portable across all major platforms.
Reliability, availability and serviceability (RAS) logs of high performance computing (HPC) resources, when closely investigated in spatial temporal dimensions, can provide invaluable information regarding system status, performance, resource utilization. These data are often generated from multiple logging systems sensors that cover many components the system. The analysis these for finding persistent insights faces two main difficulties: volume RAS makes manual inspection difficult...
We present a supercomputer-driven pipeline for in-silico drug discovery using enhanced sampling molecular dynamics (MD) and ensemble docking. also describe preliminary results obtained 23 systems involving eight protein targets of the proteome SARS CoV-2. THe MD performed is temperature replica-exchange sampling, making use massively parallel supercomputing on SUMMIT supercomputer at Oak Ridge National Laboratory, with which more than 1ms can be generated per day. have docked repurposing...
Softwarization of networked infrastructures combined with containerization codes promises unprecedented computing capabilities distributed across the federations systems and physical instruments. The development testing a software stack that implements these over an expensive production infrastructure is not cost-effective, in early stages, may potentially cause service disruptions. To address aspects, we develop Virtual Federated Science Instrument Environment (VFSIE), digital twin emulates...
Summary We measure and analyze the performance observed when running applications benchmarks before after Meltdown Spectre fixes have been applied to Cray supercomputers supporting systems at Oak Ridge Leadership Computing Facility (OLCF). Of particular interest is effect of these on selected from OLCF portfolio scale. This comprehensive study presents results experiments run Titan, Eos, Cumulus, Percival OLCF. The this are useful for HPC users serve better understand impact that two...
Recent developments in softwarization of networked infrastructures combined with containerization computing workflows promise unprecedented compute anywhere and everywhere capabilities for federations edge remote systems science instruments. The development testing software stacks that implement these over physical production federations, however, is not very practical nor cost-effective. In response, we develop a digital twin the infrastructure, called Virtual Federated Science Instrument...
Scientific workflows are increasingly being distributed across wide-area networks, and their code executions expected to span geographically dispersed computing systems. MPI has been extensively used support communications for computations, typically, over compute clusters high-performance systems within a single facility. We present case study of performance basic operations long distance connections, wherein TCP is the underlying transport. measurements execution times codes that utilize...