Research Staff 4 - Computational, High Performance Computing, Computational Science Initiative
Brookhaven National Laboratory
Computational Science Initiative
P.O. Box 5000
Upton, NY 11973-5000
Preferred Gender Pronouns (PGPs): he, him, his
Christopher is a high-performance computing specialist with over 15 years of experience developing high performance scientific software for deployment on the world's fastest supercomputers. His background is in lattice QCD — a computational branch of theoretical particle physics — with a specialization in kaon physics, and he continues to participate heavily in the field.
As a key longstanding member of the RBC & UKQCD collaborations, he leads the effort to measure direct CP-violation in kaon decays on the lattice, a phenomena that directly relates to the mystery as to why our Universe is dominated by matter with very little antimatter. He is also the lead developer for Chimbuko, a performance analysis tool for measuring and understanding the performance of HPC workflows running at-scale on supercomputers. In addition he is a founding member of the CSI Diversity, Equity and Inclusion Council.
HPC application development and physical and computational science research, including the following:
- Software design and programming - developing and designing flexible, performant and verifiable code, primarily in C++ (including language extensions such as MPI, OpenMP, CUDA, HIP and DPC++/Sycl) but with experience in Python, Perl and Bash.
- Software parallelization - extensive experience in achieving maximal parallelization in software using SIMD intrinsics, threading, GPU offloading and at the many-node level on HPC supercomputers and clusters.
- HPC systems - significant experience in deployment and running large-scale scientific production jobs on a number of supercomputer installations including Mira (and other BG/Q systems), Cori, Summit, Perlmutter and Crusher/Frontier.
- Scientific research and analysis - leading long-term scientific research programs, and data analysis including strong expertise in resampling-based statistical techniques.
- Scientific communication - scientific paper writing including a number of high-profile papers as corresponding author; presentation of work including a number of high-profile plenary conference talks, as well as seminars and colloquiums.
- (2021-present) : Primary developer of ECP-sponsored research into developing the DDHMC algorithm for accelerated lattice ensemble generation on upcoming supercomputers.
- (2020-present) : Lead developer of Chimbuko HPC performance analysis tool and member of ECP CODAR project. Supervise onboarding and assist/manage backend development team. Assist in supervision of summer interns and direction of postdocs.
- (2020-present) : Member of ECP LatticeQCD project, continuing development of K->pipi ensemble generation and measurement code to multiple GPU architectures.
- (2019-present) : Member of NESAP project for Perlmutter. Sole responsibility for porting and optimization of RBC & UKQCD K->pipi ensemble generation and measurement code to NVidia GPUs.
- (2017-present) : Lead developer of CPSfit data analysis library.
- (2017-present) : Member of Intel PathForward Exascale Computing Project. Sole responsibility for porting RBC & UKQCD K->pipi measurement code to Intel oneAPI framework in preparation for Aurora.
- (2016-2020) : Member of Intel-sponsored specialist team responsible for developing and optimizing lattice code for KNL and future Intel HPC architectures.
- (2016-2021) : Assist in supervision of Columbia University graduate student in K->pipi and related activities.
- (2014-present) : Major role in development of the Grid codebase for optimized lattice computations on a variety of target architectures.
- (2012) : Deeply involved in analysis of systematic errors on RBC & UKQCD calculation of the Delta I=3/2 K->pipi decay amplitude.
- (2011-present) : Project lead for RBC & UKQCD calculation of the Delta I=1/2 K->pipi decay amplitude, and by extension the Standard Model measure of direct CP-violation, epsilon'. Corresponding author on major PRL and PRD publications. Presented results at a number of plenary talks and seminars.
- (2011-present) : Sole responsibility for developing theory and optimized parallel code for performing lattice computations with G-parity boundary conditions. Optimized and managed ensemble generation and measurement on IBM Blue Gene/Q and Intel KNL supercomputers for Delta I=1/2 K->pipi calculation.
- (2008-2016) : Led research effort to perform simultaneous fits to multiple RBC and UKQCD lattices in order to extract basic lattice parameters. Corresponding author on two major related papers.
- Ph.D. Theoretical Physics, University of Edinburgh, 2010, (advisor: Peter Boyle)
- MPhys. 1st with Honours, The Queen's College, University of Oxford, 2007
- 10/2021 - present - Associate Staff Scientist, Computational Science Institute, BNL
- 2020 - 10/2021 - Assistant Staff Scientist, Computational Science Institute, BNL
- 2016 - 2020, Associate Staff Scientist, Columbia University (sponsored by Intel)
- 2013 - 2016, RIKEN Foreign Postdoctoral Research Fellow, RBRC, BNL
- 2010 - 2013, Postdoctoral Fellow, Columbia University
- Abbott R, Blum T, Boyle PA, et al (2020) Direct violation and the rule in decay from the standard model. Physical Review D 102:. https://doi.org/10.1103/physrevd.102.054509
- Bai Z, Blum T, Boyle PA, et al (2015) Standard Model Prediction for DirectViolation inDecay. Physical Review Letters 115:. https://doi.org/10.1103/physrevlett.115.212001
- Blum T, Boyle PA, Christ NH, et al (2016) Domain wall QCD with physical quark masses. Physical Review D 93:. https://doi.org/10.1103/physrevd.93.074505
- ISAV 2020 Best paper award for "Chimbuko: A Workflow-Level Scalable Performance Trace Analysis Tool"
- RIKEN Foreign Postdoctoral Research Fellowship (2013)
- Ken Wilson Lattice Award (2012)