• Topics of Interest • Tech. Prog. Schedule • Topic : Multi-Core • Topic : ARM Proc • Topic : Coprocessor • Topic : GPGPUs • Topic : HPC Cluster • Topic : App. Kernels • Lab. Overview • Key-Note/Invited Talks • Home




hyPACK-2013 Schedule & Tentative Technical Programme at Glance


Day 1 : October 15, 2013 (Tuesday) :

8:30 AM - 9:00 AM hyPACK-2013 Registration for FOUR Days Prog.

9:00 AM - 9:30 AM Welcome & Inauguration Ceremony  
More.......download hyPACK-2013 Inauguration

More.......download hyPACK-2013 Overview

Dr. S.K.Udgata;   Director, CMSD, University of Hyderabad
Dr.VCV.Rao;   hyPACK-2013 Workshop Co-ordinator, C-DAC-Pune

9:30 AM - 10:00 AM An Overview of hyPACK-2013
Technical Prog. (Class-Room Lectures, Academia /Industry Talks)
An Overview of Lab. Sessions : HPC Cluster with Intel Xeon Phi
  Coprocessors & GPGPUs/GPU Computing; Power-aware Computing
An Introduction to Intel Xeon Phi Coprocessors - Prog. Approach
Speaker : C-DAC
10:00 AM - 10:30 AM Coffee & Tea Break
10:30 AM - 11:15 AM Classroom Lecture & Laboratory Session
- Topic : An Overview of Intel Xeon Phi Programming - Offload Pragmas & Performance Issues
Speaker : C-DAC

11:15 AM - 11:45 AM Classroom Lecture & Laboratory Session
Topic : An Overview of Intel Xeon-Phi Coprocessors - System Software & Compiler Optmisation - Vectorization and Hands-on Session
Speaker : C-DAC

11:45 AM - 12:45 PM Classroom Lecture & Demonstration:
Topic : Intel Xeon-Phi Architecture - Prog. Models - Compilation features; Compiler tips, Compiler Vectorization reports; Compiler Directives & Demonstration Benchmarks - OpenMP on Xeon Host and Xeon-Phi Coprocessor

12:45 PM - 1:00 PM Lab. Session:
Topic : Hands-on session on Intel Xeon-Phi Coprocessor - OpenMP framework on Intel Xeon-Phi - Tuning & Performance
Instructor : C-DAC
1:00 PM -1:45 PM Lunch Break
1:45 PM - 2:30 PM Laboratory Session
Instructors : C-DAC & Intel
Programming - Intel Xeon-Phi Compiler Offload Pragmas; Compiler Technology & Vectorization - Numerical Computations
2:30 PM - 4:00 PM Demonstration & Lab. Session:
Instructors : C-DAC & Intel
Programming : Mixed Programming (OpenMP, Intel TBB, Pthreads); Compilation & Vectorization - Numerical Computations based on Intel MKL on Xeon Phi Coprocessors; Tuning and Performance of Benchmarks on Intel Xeon-Phi;Compilation - Intel Offload pragmas
4:00 PM - 4:15 PM Coffee Break
4:15 PM - 5:00 PM Keynote Talk (Academic) :
Numerical Linear Algebra (NLA); Intel MKL - Performance on Intel Xeon-Phi Coprocessors
Speaker : C-DAC

5:00 PM - 5:45 PM Lab. Session:
Instructors : C-DAC
Intel Xeon-Phi Prog (OpenMP, TBB, Pthreads); Performance of Benchmarks - Numerical Computations: Compiler Optimizations & Vectorization; Intel MKL Math Kernels on Xeon Phi; Thread Affinity - OpenMP & Pthreads on Intel Xeon Phi; Performance of Application Kernels on Intel Xeon-Phi
5:45 PM - 6:30 PM Invited Talk :
An Overview of MPI programming - Intel Xeon-Phi Coprocessors & Performance Issues

Day 2 : October 16, 2013 (Wednesday) :

9:00 AM - 9:45 AM Classroom Lecture & Laboratory Session
Topic : Prog. Intel Xeon Phi - Prog. Affinity Concepts Intel Xeon-Phi Programming Models - Intel TBB, Overview of Co-Processor Prog. Models; Tuning & Performance of NLA Kernels

9:45 AM - 10:30 AM Classroom Lecture & Laboratory Session
Topic : Prog. Intel Xeon Phi -Compiler Options; Cilk Plus Compiler tips, Compiler Vectorization reports; Compiler Directives - Memory alignment; Math Kernel Library - Performance Results
10:30 AM - 10:45 AM Coffee & Tea Break
10:45 AM - 11:15 AM Classroom Lecture & Laboratory Session
Topic : MPI execution models on Intel Xeon and Intel Xeon Phi coprocessors, including pure MPI or hybrid MPI applications

11:15 AM - 12:15 PM Keynote Talk (Industry) :
Topic : An Overview of Intel Xeon-Phi Co-processor Archi.; An Overview of Co-Processor System Software; Compiler Offload Pragmas & Tuning for performance on Xeon Phi(tm) coprocessor using VTune Amplifier XE
Speaker : Intel
12:15 PM - 1:00 PM Keynote Talk (Industry) :
Topic : Intel Xeon Phi / Demos: Compilers, VTune Amplifier Demonstrate Compilers and VTune AXE on Intel ; Tuning methodologies; Intel(r) Trace Analyzer and Collector (ITAC) on MICs
Speaker : Intel
1:00 PM -1:45 PM Lunch Break
1:45 PM - 3:00 PM Demonstration & Lab. Session:
Instructors : C-DAC
MPI execution models on Intel Xeon processors and Intel Xeon Phi coprocessors, including pure MPI or hybrid MPI applications.; Mixed Prog. (MPI-OpenMP, MPI-TBB, OpenCL, Pthreads); Basic Programs based on Intel Xeon-Phi Co-processors; Bandwidth Calculation Matrix Computations; Using Intel Xeon-Phi Tools for Numerical Computations
3:00 PM - 4:00 PM Keynote Talk (Industry) :
Performance of Application Kernel - based on OpenMP /MPI - Intel Xeon-Programming Framework
Speaker : Intel
4:00 PM - 4:15 PM Coffee Break
4:15 PM - 5:15 PM Lab. Session:
Topic : Mixed Prog. (MPI-OpenMP, MPI-TBB, MPI-Pthreads) on Intel Xeon Co-Processors; I/O files on Intel Xeon Co-processors; Memory Map (mmap) Examples on Intel Xeon-Phi Co-processors; MPI Prog. for Numerical Computations on Intel Xeon Phi; Intel Tool Kit Demonstration
Instructor : C-DAC

5:15 PM - 6:00 PM Classroom Lecture & Lab (part-II):
Prog. Intel Xeon Phi - Memory Map (mmap) & Huge Page Enabling
Instructors : C-DAC


Day 3: October 17, 2013 (Thursday) :

9:00 AM - 9:45 AM Classroom Lecture & Laboratory Session
Topic : Introduction to GPU Computing : Memory Optimization; Tuning & Performance on CUDA enabled NVIDIA GPUs
Instructors : C-DAC
9:45 AM - 10:30 AM Classroom Lecture & Laboratory Session
Topic : CUDA - enabled NVIDIA GPUs - Use of CUDA Toolkit Math Libraries - OpenACC Pragmas Framework - CUDA Enabled NVIDIA Multi-GPUs

10:30 AM - 10:45 AM Coffee & Tea Break
10:45 AM - 11:45 AM Keynote Talk (Industry) :
Topic : An Overview of CUDA enabled NVIDIA GPUs - Programming & Performance Issues
Speaker : NVIDIA

11:45 AM - 12:45 PM Keynote talk (Industry )
Topic : An Overview of CUDA enabled NVIDIA GPUs - Programming & Performance Issues.
Speaker : NVIDIA
1:00 PM - 2:00 PM Lunch Break
2:00 PM - 3:00 PM Laboratory Session & Demonstration
Instructors : C-DAC
Lab. Session: An overview of CUDA enabled NVIDIA GPUs / OpenCL - GPGPUs/Example Programs
3:00 P M - 4:00 PM Laboratory Session & Demonstration
Topic : Tuning & Performance on CUDA enabled NVIDIA-GPUs; Matrix-matrix multiplication - tiled techniques for partitioning of a matrix, shared memory optimization, Warp level parallelism; Tuning & Performance on Multi-CUDA enabled NVIDIA-GPUs

4:00 PM - 4:15 PM Coffee Break
4:15 PM - 5:00 PM Keynote Talk (Academic):
Topic : Tuning & Performance on CUDA NVIDIA GPUs
5:00 PM - 5:30 PM Instructors : C-DAC
Laboratory Session
Application kernels based on Mixed Prog. (MPI-CUDA, Pthreads-CUDA & OpenMP-CUDA); CUDA SDK ToolKit Demonstration; Prog.On Heterogeneous Comp. Platforms - AMD-APP; Basic OpenCL Programs based on Single /Multiple GPUs on AMD-APP GPUs; Use of Work Groups & Work-items - Memory Optimizations; Prog. on ARM Multi-Cores with CUDA NVIDIA carma - Using NVML APIs
5:30 PM - 6:00 PM Instructors : C-DAC
Laboratory Session
Measurement of Power Consumption - Performance of Application Kernels - Using NVML Lib Calls


Day 4: October 18, 2013 (Fridday) :

9:00 AM - 9:45 AM Classroom Lecture & Laboratory Session
Topic : Heterogeneous Programming - CUDA enabled NVIDIA GPUs /AMD APP - OpenCL; Tuning & Performance - Matrix Computations; AMD APP Tech. - SDK & Prog. Env /Libraries
9:45 AM - 10:30 AM Classroom Lecture & Lab :
Topic : An Overview of HPC GPU Cluster - OpenCL Performance Issues - Numerical Linear Algebra; AMD APP Tech - Tuning & Performance OpenCL; Demonstration of Application Kernels
Speaker :

10:30 AM - 10:45- AM Coffee & Tea Break
10:45 AM - 11:45 AM Keynote Talk (Industry):
Topic : Measurement of Power Consumption - Systems with Coprocessors/Accelerators; Case Studies - Partial differential Eqs - Solution of Matrix Systems on Cluster with Single /Multiple GPUs
11:45 AM - 12:45 PM Keynote Talk (Academic):
Topic : Performance of Application Kernels on Parallel Processing Platforms with GPU Accelerators
Speaker : Samrit Maity, C-DAC, Pune

1:00 PM - 1:45 PM Lunch Break

2:00 PM - 3:00 PM Laboratory Session
Instructors : C-DAC
Hands-on session on NVIDIA GPUs /AMD OpenCL;Prog. on ARM Multi-Core system with CUDA NVIDIA carma - Using NVML APIs
3:00 PM - 4:00 PM Lab. Session - Demonstration:
Topic : Programming based on OpenCL, Tuning & Performance of OpenCL on GPGPUs; matrix-matrix multiplication - algorithms based on OpenCL; Shared memory optimization, Wavefront level parallelism; - Memory Optimizations

4:00 PM - 4:15 PM Coffee Break
4:15 PM - 5:00 PM Keynote Talk (Academic):
Topic : An Overview of Application Kernels on Parallel Processing Systems with Multi-GPU - Power aware Performance Issues - NVML Library calls & external Power-Off Meter
5:00 PM - 5:45 PM Lab. Session:
Topic : Example programs on host-cpu (Pthreads, MPI, OpenMP) and OpenCL on Multiple GPUs; Tuning & Performance of Matrix Computations on AMD-APPs; Memory Optimization on AMD APP - OpenCL; Application kernels based on Mixed Prog. (MPI,Pthreads, OpenMP- with OpenCL); OpenCL programs for Numerical Linear Algebra on HPC GPU Cluster (OpenCL on NVIDIA/AMD-APP GPUs) Benchmarks
5:45 PM - 6:00 PM - Closure of hypack-2013
Centre for Development of Advanced Computing