Kevin Bowers

Kevin Bowers Email and Phone Number

High Performance and Scientific Computing Research and Development @ Jump Trading LLC
Kevin Bowers's Location
Arlington Heights, Illinois, United States, United States
Kevin Bowers's Contact Details
About Kevin Bowers

- Development and application of advanced theoretical and computational techniques to enabling technologies; enjoy adapting physics, electrical engineering, mathematics and computer science methods to challenging problems in any discipline- Proven ability to do innovative R&D rapidly even in new fields with experience in a broad spectrum of organizations (large to small to independent, public to private to academic, biotech to defense to telecom to semiconductors to finance)- Proven ability to design breakthrough algorithms and to optimize (both algorithmically and in implementation) software, hardware, wired networks and wireless networks for ultra high performance; computational experience on platforms ranging from desktops to distributed supercomputers with hundreds of thousands of cores producing data at rates of hundreds of terabytes every second.Specialties: Scientific Computing: Particle and electromagnetic simulation (Boltzmann and Maxwell equations), PDEs (initial value, boundary value and eigenvalue systems), matrix algorithms (dense, structured and sparse), pattern recognition (hidden Markov, clustering), signal processing (FFT, wavelets, filtering, compression, digital modulation), C, C++, assembly, MATLAB/MEX, LaTeX, parallel programming (distributed memory/MPI, shared memory/pthreads, data parallel/SIMD), Linux ...

Kevin Bowers's Current Company Details
Jump Trading LLC

Jump Trading Llc

View
High Performance and Scientific Computing Research and Development
Kevin Bowers Work Experience Details
  • Jump Trading Llc
    Chief Science Officer / Head Of Research And Development
    Jump Trading Llc 2011 - Present
    Chicago, Illinois, Us
    Researched and developed algorithms, software and hardware for low latency trading and for massively parallel optimization of quantitative trading strategies (work covered the entire trading technology stack from low level physical layer through high level trading APIs).
  • Jump Trading Llc
    Systems Architect
    Jump Trading Llc 2010 - 2011
    Chicago, Illinois, Us
    See above
  • Los Alamos National Lab
    Guest Scientist
    Los Alamos National Lab 2004 - 2014
    Los Alamos, Nm, Us
    Architected software for petaflop (million billion floating point operations per second) scale supercomputers:- VPIC (below) was a flagship code for the world's first petaflop capable supercomputer (Roadrunner, 13K AMD Opteron cores and 13K IBM Cell chips); the initial port sustained roughly half a petaflop during Roadrunner acceptance and assessment trials. VPIC continues to be used extensively in highly cited research and remains an anchor code for more recent systems, including planned exascale (billion billion floating point operations per second) supercomputers.- Received multiple Los Alamos Distinguished Performance Awards and Defense Programs Awards of Excellence, was a finalist for the Supercomputing 2008 Gordon Bell Prize and work is publicly showcased at the Bradbury Science Museum (exhibit "Roadrunner: Advanced Architecture for Algorithms and Applications").
  • Los Alamos National Lab
    Technical Staff Member
    Los Alamos National Lab 2002 - 2004
    Los Alamos, Nm, Us
    Applied Physics Division (X)Plasma Physics Group (X-1)- Developed algorithms for parallel supercomputers, culminating in VPIC, an ultra high performance first-principles kinetic particle-in-cell code (on a per node basis, usually faster than comparable codes by one to two orders of magnitude; used for modeling plasma instabilities / turbulence, laser-plasma interactions, relativistic beams, magnetic reconnection ...)- Applied advanced analysis techniques to multi-terabyte multi-dimensional data sets (implemented maximum-likelihood and maximum-entropy PDF estimation methods to couple and analyze PIC, Monte-Carlo and hydrodynamic simulations; invoked spectral, time-frequency, time-scale and wavelet methods to analyze plasma turbulence)- Modeled relativistic electron beam X-ray radiographic sources
  • D.E. Shaw Research Llc
    Scientist
    D.E. Shaw Research Llc 2004 - 2010
    New York, New York, Us
    - Researched scientific computing: Novel parallel decompositions and distributed cache protocols; communication, thread, SIMD and NUMA optimization (latency optimized network protocols, parallel FFTs, SSE/AltiVec code generation); robust (even bit-level reversible!) integration/thermodynamic sampling with single, double and custom precision; patents- Architected a massively parallel molecular dynamics code (Desmond) based on this: designed code, implemented core, integrated work of several developers; more numerically robust, more scalable and/or faster than comparable codes---won best paper at Supercomputing 2006, featured in Nature, successfully commercialized- Assisted with design and bring up of large clusters (thousands of cores) dedicated to Desmond including performance modeling / competitive analysis of HPC technology (heterogeneous multicore CPUs, GPUs)- Contributed to design of a custom ASIC for MD (Anton): parallelization, numerical formats, instruction set design; Anton won best paper and Gordon Bell prize at Supercomputing 2009 and the Test of Time Award at Supercomputing 2023.
  • Lucent Bell Labs / Agere Systems
    Member Technical Staff
    Lucent Bell Labs / Agere Systems 2001 - 2002
    Physical Sciences Research DivisionOpto-Electro-Mechanical Integration Research and Optical MEMS Groups- Applied the PROPHET and IMSIL codes to optimize SiGe dopant implantation processes- Developed models of dopant diffusion in alloy lattices- Derived analytic models of electro-mechanical coupling in MEMS mirror arrays- Implemented a highly efficient 2.5d frequency domain electromagnetic scattering code (intended for optimization of waveguide devices and diffractive / binary optics, optimized matrix algorithms achieve theoretical memory limited performance on SMP workstations, automatic boundary mesh generation able to handle arbitrary linear materials, unbounded systems and systems with wide scale variation, compact representation of scattered fields achieved through novel application of Huygens' principle)
  • Los Alamos National Lab
    Graduate Research Assistant
    Los Alamos National Lab 1997 - 1997
    Los Alamos, Nm, Us
    Computing, Information and Communication Division (CIC)Image Processing and Analysis Group (CIC-3)- Researched multilingual handwritten document analysis (developed a medial axis / skeletonization / thinning algorithm; implemented a C++ class for training and evaluation of hidden Markov models; implemented code to extract and cluster feature vectors from document images; developed and evaluated of several algorithms to perform language identification on handwritten documents---the top algorithm correctly classified out of eight choices the language with 88% accuracy, top two choices over 97%)
  • Intel Corporation
    Intern
    Intel Corporation 1996 - 1996
    Santa Clara, California, Us
    i960 Embedded RISC Processor Component and Tools Applications Engineering - Assisted OEMs with Intel component related design issues---dealt with ~700 issues- Identified sales opportunities and recommended components to OEMs for new designs- Qualified potential errata in i960 software tools, components and manuals- Trained other Intel employees about the i960 processor and applications- Planned, designed, and completed technical projects to demonstrate the use and suitability of Intel components in designs (benchmarked the i960 family in various embedded applications, wrote a bootstrap loader for the i960CF on a Cyclone development board, developed ethernet drivers for an i82596 MAC to i960CF hardware interface, developed a software interface between a Pentium desktop and an i960Hx PCI-SDK board, created a preemptive multitasking kernel for i960Hx processor for the i960 Programmer's Cookbook)

Kevin Bowers Skills

High Performance Computing Algorithms Simulations Scientific Computing C Physics Parallel Computing C++ Latex Mpi Programming Theory Parallel Programming Numerical Analysis Signal Processing Data Structures Science Linux Distributed Systems Unix Mathematics Computer Science Simulation Matlab Mathematical Modeling Computer Architecture Fortran X86 Assembly Software Development Experimentation Clustering Image Processing Optimization R&d Software Engineering Applied Mathematics Software Design Statistics Optimizations Numerical Simulation Computational Physics Monte Carlo Simulation Embedded Systems Dsp Optics Filtering Pattern Recognition Machine Learning Cuda

Kevin Bowers Education Details

  • University Of California, Berkeley
    University Of California, Berkeley
    Electrical Engineering And Computer Science
  • Purdue University
    Purdue University
    Electrical Engineering
  • The University Of Akron
    The University Of Akron
    Mathematics

Frequently Asked Questions about Kevin Bowers

What company does Kevin Bowers work for?

Kevin Bowers works for Jump Trading Llc

What is Kevin Bowers's role at the current company?

Kevin Bowers's current role is High Performance and Scientific Computing Research and Development.

What is Kevin Bowers's email address?

Kevin Bowers's email address is ke****@****ail.com

What schools did Kevin Bowers attend?

Kevin Bowers attended University Of California, Berkeley, Purdue University, The University Of Akron.

What are some of Kevin Bowers's interests?

Kevin Bowers has interest in Mathematics, Classical Music, New Technology, Table Tennis, Law, Politics, Computer Graphics, Chess, Audiophile, Jazz Music.

What skills is Kevin Bowers known for?

Kevin Bowers has skills like High Performance Computing, Algorithms, Simulations, Scientific Computing, C, Physics, Parallel Computing, C++, Latex, Mpi, Programming, Theory.

Free Chrome Extension

Find emails, phones & company data instantly

Find verified emails from LinkedIn profiles
Get direct phone numbers & mobile contacts
Access company data & employee information
Works directly on LinkedIn - no copy/paste needed
Get Chrome Extension - Free

Aero Online

Your AI prospecting assistant

Download 750 million emails and 100 million phone numbers

Access emails and phone numbers of over 750 million business users. Instantly download verified profiles using 20+ filters, including location, job title, company, function, and industry.