Rajib kumar nath



Download 53.17 Kb.
Date09.06.2018
Size53.17 Kb.
#53770
RAJIB KUMAR NATH

9450 Gillman Drive #80361 • La Jolla• CA 92092



rknath@cs.ucsd.edu• 865-242-9675 • www.cse.ucsd.edu/~rknath

Objective

I am looking for an internship in the area of energy efficient Hybrid Memory Architecture and Operating System


Research Interest

Energy Efficient System, Computer Architecture, Hybrid Memory, Operating System, and Data Centre


Education

PhD in Computer Science September, 2010 – Present

University of California, San Diego (UCSD), CA, USA
Masters of Science in Computer Science and Computational Science August 2010

University of Tennessee, Knoxville (UTK), TN, USA
Bachelor of Science in Computer Science and Engineering June 2005

Bangladesh University of Engineering and Technology (BUET), Dhaka, Bangladesh
Professional Experience

Summer Intern May 2009 – August 2009

Math Kernel Library, Intel(R), Hillsboro, OR, USA­­­­­­

Mentor: Dr. Greg Henry

Area: Compiler and Programming Language

  • Created a new generation of a generic assembly code optimizer tool called Super Compiler

  • Solved the timing problem that was inherent in all previous generations of the tool

  • Resulted in Super Compiler generated code integration within current release of Math Kernel Library


Summer Research Intern May 2008 – August 2008

Paris Project Team, French National Institute for Research in Computer Science and Control (INRIA), Rennes, France

Mentors: Dr. Christine Morin & Dr. Thomas Ropars

Area: Grid Computing and Fault Tolerance

  • Proposed a Fault Tolerance System for an existing Grid Middleware, VIGNE

  • Implemented the consensus protocol and a weak failure detector for VIGNE


Quantitative Software Engineer August 2005 – August 2007

Stochastic Logic Limited, ACI Ltd, Dhaka, Bangladesh

Mentor: Dr. Arif Dowla

Area: Computational Finance

  • Built time series modeling and option pricing tools using Matlab and S-plus

  • Constructed a module for Time Series smoothing using non parametric Kernel Regression

  • Developed a heuristic based approach for Regime Switch Detection in Financial Time Series


Research Experience

Graduate Research Assistant September 2010 – Present

System Energy Efficiency Laboratory, University of California, San Diego, CA, USA

Mentor: Dr. Tajana Rosing

Research Area: Hybrid Memory and Energy Efficiency, Computer Architecture, Operating System

  • Studying scalable, high performance, energy efficient and reliable hybrid memory architecture


Master’s Thesis: Accelerating Dense Linear Algebra for GPUs, Multicores and Hybrid Architectures August 2010

Department of Computer Science, University of Tennessee, Knoxville, TN, USA

Mentor: Dr. Jack Dongarra & Dr. Stanimire Tomov
Graduate Research Assistant August 2008 – August 2010

Innovative Computing Laboratory, University of Tennessee, Knoxville, TN, USA

Project: Matrix Algebra on GPU and Multicore Architectures (MAGMA)

Mentors: Dr. Jack Dongarra & Dr. Stanimire Tomov

Research Area: High Performance Computing

  • Developed a set of highly optimized Basic Linear Algebra Subprograms(BLAS) for MAGMA

  • Devised an auto tuning tool that defines different tuning parameters and searches for near optimal values for GPU kernels; consistently outperformed contemporary NVIDIA CUDA BLAS

  • Formulated a simple but innovative Pointer Redirecting approach for GPU kernels that provides consistent high performance in a wide range of problem sizes

  • Developed matrix-matrix multiplication code for Fermi which is included as part of CUBLAS-3.2


Project: Parallel Linear Algebra for Scalable Multi-core Architectures (PLASMA)

Mentors: Dr. Jack Dongarra, Dr. Jakub Kurzak, & Dr. Emmanuel Agullo

Research Area: High Performance Computing

  • Optimized single precision floating point operations for Intel x86_64 architectures

  • Implemented an auto tuning component for a multicore dense linear algebra library


Graduate Research Assistant January 2008 – May 2008

Department of Computer Science, University of Tennessee, Knoxville, TN

Mentor: Dr. Michael Langston

Research Area: Algorithm Complexity

  • Analyzed and compared different clustering algorithm for a well-known yeast data


Undergraduate Thesis: Load Balancing in a Busy Web Server June 2004 – June 2005

Department of Computer Science and Engineering, BUET, Dhaka, Bangladesh

Mentor: Dr. Shamsul Alam

Research Area: Networking and Distributed System

  • Proposed and simulated a metrics based load balancing technique


Participant

Gene Golub Siam Summer School on Numerical Linear Algebra, SIAM, Brindisi, Italy June, 2010

Best Practice in Teaching Workshop, University of Tennessee, Knoxville, TN, USA August – December, 2009
Teaching Experience

Graduate Teaching Assistant August 2007– May 2008

Department of Computer Science, University of Tennessee, Knoxville, TN, USA
Lecturer June 2005 – August 2005

Computer Science Department, United International University, Dhaka, Bangladesh
Selected Course Projects:

Human Activity Recognition using Hidden Markov Model and Recurrent Neural Network, UTK 2009

N-Body Simulation to model Solid to Liquid state transition using OpenGL, UTK 2009

Simultaneous Adversarial Multi Agent Game Learning in Grid Soccer Domain, UTK 2008

Multi Robot Predator-Prey in Player Stage Simulator, UTK 2008

Survey of Loop Optimization Techniques, UTK 2008

Multi User peer-to-peer Fault Tolerant Distributed Chat, UTK 2007

Wireless Peer-to-Peer Data Transmission, BUET 2004

OSI Layer Implementation using Java, BUET 2004

Compiler for a simple CFG using C, BUET 2004

Multithreaded HTTP Server using Java, BUET 2003
Publications:

H. Ltaief, S. Tomov, R. Nath, P. Du, J. Dongarra, A Scalable High Performant Cholesky Factorization for Hybrid “Multicore & MultiGPU” System, VECPAR 2010, CA, USA.


S. Tomov, R. Nath, H. Ltaief, J. Dongarra, Dense Linear Algebra Solvers for Multicore with GPU Accelerators, HIPS 2010, held in conjunction with IPDPS 2010, GA, USA.


R. Nath, S. Tomov, J. Dongarra, Book Chapter, In Scientific Computing with Multicore and Accelerators, Computational Science Series, Chapman & Hall/CRC, 2010

R. Nath, S. Tomov, J. Dongarra, Accelerating GPU Kernels for Dense Linear Algebra, VECPAR 2010, CA, USA.

S. Tomov, R. Nath, P. Du, J. Dongarra, MAGMA Version 0.2 users’ Guide, Innovative Computing Laboratory,

University of Tennessee, Knoxville, TN, 2009


R. Nath, S. Tomov, J. Dongarra, An Improved Magma Gemm For Fermi Graphics Processing Units, The International

Journal of High Performance Computing Applications, 2010.

S. Tomov, R. Nath, J. Dongarra, Accelerating the reduction to upper Hessenberg, tridiagonal, and bidiagonal forms

through hybrid GPU-based computing Submitted, PARCO 2011.

J. Kurzak, R. Nath, P. Du, J. Dongarra, An Implementation of the Tile QR Factorization for a GPU and Multiple CPUs,

PARA' 2010, Iceland.

E. Agullo, J. Dongarra, R. Nath, S. Tomov, Autotuned dense QR factorization for multicore architectures, Technical

Report, Innovative Computing Laboratory, TN, USA, 2010


Posters:

E. Agullo, C. Augonnet, J. Dongarra, H. Ltaief, R. Namyst, R. Nath, J. Roman, S. Thibault, S. Tomov, Scheduling

Cholesky Factorization on Multicore Architectures with GPU Accelerators, SAAHPC 2010, TN, USA.

R. Nath, J. Dongarra, S. Tomov, H. Ltaief, and P. Du, Recent Developments in the MAGMA, ATIP, SC 2009, OR, USA.

E. Agullo, B. Hadri, H.Ltaief, S. Tomov, J. Demmel, J. Kurzak, P. Luszczek, V. Volkov, J. Dongarra, J. Langou, R. Nath, A.

YarKhan, Numerical Linear Algebra on Emerging Architecture, SC 2009, OR, USA.
Computer Skills:


  • Languages: C, C++, Java, Assembly Language, NVIDIA-CUDA, Prolog, Python, PHP, HTML, Perl, shell

  • Operating Systems: Linux, Windows, Mac-OS

  • Database: Oracle, MySQL

  • Mathematical Tools: S+, Matlab, Mathematica, BLAS, MKL, PLASMA, MAGMA, LAPACK

  • Parallel Computing Tools: MPI, posix thread, OpenMp

  • Performance Tools: PAPI, TAU, JumpShot

  • Others: LATEX , Player-Stage, HMM toolkit HTK

  • Simulation Tools: M5Simulator


Awards

Graduate Fellowship, University of California, San Diego, USA 2010

Travel Grant , National Science foundation, USA 2009

Accommodation Grant, Society of industrial and Applied mathematics, USA 2009

CUDA Center of Excellence (CCOE) to MAGMA Project, Nvidia, USA 2009

Travel Grant , Asian Technology Insurance Program, USA 2009

Graduate Assistantship, University of Tennessee, Knoxville, TN, USA 2007–2010

Dean’s Scholarship & Merit Scholarship, BUET, Dhaka, Bangladesh 2000, 2003–2004

National Education Board Scholarship, Board of Education, Dhaka, Bangladesh 1997–2005
Volunteer Work

Student Volunteer, Supercomputing, Portland, OR 2009

Blood Donor, Medic Regional Blood Centre, Knoxville, TN 2009 – 2010

Blood Donor, Badhon: a voluntary blood donor organization, Dhaka, Bangladesh 2000 – 2007

Volunteer, Students for Flood Victims of Bangladesh, BUET, Dhaka, Bangladesh 2003
Organizations

Hridoya Bangladesh 2007 – Present

ACM 2008 – Present
Talks

R. Nath, Accelerating Dense Linear Algebra for GPUs, Vecpar 2010, UC Berkeley, June, 2010

R. Nath, Autotuned Dense Linear Algebra Library, Mathematic, University of Tennessee, Knoxville, December, 2009

R. Nath, Inferring Human Intention in Peer-to-Peer Human Robot Teaming for Shared Workspace



Interaction, Distributed Intelligence Laboratory, University of Tennessee, Knoxville, November, 2009

R. Nath, Supercompiler: a Performance Tuning Tool, Intel(R), August, 2009



R. Nath, Fault Tolerance in VIGNE, INRIA, Rennes, France, August 2008
Download 53.17 Kb.

Share with your friends:




The database is protected by copyright ©ininet.org 2024
send message

    Main page