Skip to navigation Skip to content
Careers | Phone Book | A - Z Index
Performance and Algorithms Research
Genomev1

Exabiome Project

Exabiome project is developing exascale computing tools to solve previously infeasible science problems in genomic analysis.

RAPIDS Thumbnail 2

RAPIDS2

RAPIDS2 SciDAC5 Institute assists Office of Science application teams in overcoming computer science, data, and AI challenges in using DOE supercomputing resources to achieve scientific breakthroughs.

Endurable illo thumbnail

ENDURABLE

The ENDURABLE project aims to provide the scientific community with tools to aggregate data robustly and train our deep learning models.

graph 5727061 1280

Scalable Graph Learning for Scientific Discovery

Graph representation learning is transforming scientific domains from structural biology to particle physics, transportation, and beyond.

020919 bes nanoparticles

SciDAC-5: DECODE

Data-driven Exascale Control of Optically Driven Excitations in Chemical and Material Systems (DECODE) combines exascale and non-traditional machine learning for advances in optical systems.


The Performance and Algorithms Research Group focuses on the research and development of technologies and algorithms that enhance the performance, scalability, and energy efficiency of applications running on the Department of Energy's multicore-, manycore-, and accelerator-based supercomputers.  Moreover, we develop performance models to understand the inherent bottlenecks in today's systems as well as predict the performance and bottlenecks of tomorrow's exascale systems.  To that end, we have formed strong research collaborations with computer science, computer architecture, applied math, and application research teams.

Group Leader: Lenny Oliker

Research

ExaBiome

Exabiome project is developing exascale computing tools to solve previously infeasible science problems in genomic analysis.

Parallel Primitives for Randomized Algorithms on Sparse Data

Parallel Primitives for Randomized Algorithms on Sparse Data project targets scalable randomized methods broadly within the context of scientific data analysis.

SciDAC-5 SAP: DECODE

This project aims to harness exascale computing and non-conventional machine-learning approaches in order to design tailored optical excitations for controlling electron-driven dynamics in chemical/material systems using real-time time-dependent density functional theory (RT-TDDFT).

SciDAC-5 SAP: Correlated Electrons in QM

This project aims at understanding, describing, and predicting the dynamics of the individual electronic states under non-equilibrium (NE) conditions.

Scalable Graph Learning for Scientific Discovery

Scalable Graph Learning for Scientific Discovery targets graph representation learning (GRL). GRL is transforming scientific domains like structural biology, computational chemistry, particle physics, transportation, and program analysis.

Exagraph

ExaGraph Co-Design Center of the Department of Energy Exascale Computing Project (ECP) uses combinatorial kernels, with key examples being smart power grid, computational biology, computational chemistry, wind energy, and national security.

RAPIDS2

RAPIDS2 SciDAC5 Institute assists Office of Science (SC) application teams in overcoming computer science, data, and AI challenges in the use of DOE supercomputing resources to achieve scientific breakthroughs.

ENDURABLE

The ENDURABLE project aims to provide the scientific community with tools to aggregate data robustly and train our deep learning models.

Performance Analysis of AI Hardware and Software

Understanding the interplay between science, AI method, framework, and architecture is essential to quantifying the computational potential for current and future architectures running AI models and for identifying the bottlenecks and the ultimate limits of today's models.

SciDAC-5 Institute

PAR researchers are engaged in performance modeling, machine learning, communication runtimes, and performance optimization research for applications in the Scientific Discovery through Advanced Computing (SciDAC) initiative.

ECP PROTEAS-TUNE

ROTEAS-TUNE is a multi-institutional ECP software technology project spanning the topics of compilers, code generation, auto-tuning, and profiling.

Roofline Performance Model

Roofline is a visually intuitive performance model used to bound the performance of various numerical methods and operations running on multicore, manycore, or accelerator processor architectures.

High Performance Geometric Multigrid

High Performance Geometric Multigrid (HPGMG-FV) is a benchmark designed to proxy the finite volume based geometric multigrid linear solvers found in adaptive mesh refinement (AMR) based applications.

Previous Projects

Over the last 10 years, researchers in the Performance and Algorithms research group have led a number of research projects spanning performance optimization, performance modeling, co-design, supercomputer benchmarking, and application of novel algorithms.

More from Research »