Presentation is loading. Please wait.

Presentation is loading. Please wait.

MATRIX MULTIPLY WITH DRYAD B649 Course Project Introduction.

Similar presentations


Presentation on theme: "MATRIX MULTIPLY WITH DRYAD B649 Course Project Introduction."— Presentation transcript:

1 MATRIX MULTIPLY WITH DRYAD B649 Course Project Introduction

2 Matrix Multiplication Fundamental kernel algorithm used by many applications Examples: Graph Theory, Physics, Electronics

3 Scalability Issues: Run on single machine: Memory overhead increase in terms of N^2 CPU overhead increase in terms of N^3 Run on multiple machines: Communication overhead increase in terms of N^2

4 Matrix Multiply Approaches Programming Mdoel AlgorithmCustomized Libraries User Implementation SequentialNaïve approach, tiles matrix multiply, Blas_dgemm Vendor supplied package (ie, Intel, AMD Blas), ATLAS Fortran, C, C++, C#, Java Shared memory parallelism Row PartitionATLASMulti Threads, TPL, PLINQ, OpenMP Distributed memory parallelism Row Column Partition, Fox Algorithm ScalePackOpenMPI, Twister, Dryad

5 Why DryadLINQ? Dryad is a general purpose runtime that supports the processing of data intensive application in Windows DryadLINQ is a high level programming language and compiler for Dryad Applicability: Dryad transparently deal with the parallelism, scheduling, fault. tolerance, messaging, and workload balancing issues. SQL-like interface, based on.NET platform, easy to have code. Performance: Intelligent job execution engine, optimized execution plan. Scale out for thousands of machines.

6 Parallel Algorithms for Matrix Multiplication MM algorithms can deal with matrices distributed on rectangular grids No single algorithm always achieves best performance on different matrix and grid shapes. MM Algorithms can be classified into categories according to the communication primitives Row Partition Row Column Partition Fox Algorithm (BMR) – broadcast, multiply, roll up

7 Row Partition Heavy communication overhead Large memory usage per node The full Matrix B is copied to every node The Matrix A row blocks are distributed to each node Pseudo Code sample: Partition matrix A by rows Broadcast matrix B Distributed matrix A row blocks Compute the matrix C row blocks

8 Row Column Partition Heavy communication overhead Scheduling overhead for each iteration Moderate memory usage Pseudo Code sample: Partitioned matrix A by rows Partitioned matrix B by columns For each iteration i: broadcast matrix A row block i distributed matrix B column blocks compute matrix C row blocks

9 Fox Algorithm Stage OneStage Two

10 Fox algorithm Less communication overhead than other approach Scale well for large matrices sizes Pseudo Code sample: Partitioned matrix A, B to blocks For each iteration i: 1) broadcast matrix A block (i%N, i%N) to row i 2) compute matrix C blocks add the to the previous result 3) roll-up matrix B block

11 Performance Analysis on Fox algorithm Cache Issue Cache miss (size), pollution, confliction Tiles matrix multiply Memory Issue Size (memory paging) Bandwidth, latency Cache Size Turning Point Absolute performance degrade as problem size increase for both cases Single node performance worse than multiple nodes due to memory issue.

12 Multicore level parallelism To use every core on a compute node for Dryad Job, the task must be programmed with multicore technology. (i.e. Task Parallel Library, Thread, PLINQ) For each thread, it will compute one row in matrix C or several rows in matrix C depends on the implementation. By using TPL or PLINQ, the optimization for threads is implicit and easier to use.

13 Timeline for term long project Stage One Familiar with HPC cluster Sequential MM with C# Multithreaded MM with C# Performance comparison of above two approaches Stage Two Familiar with DryadLINQ Interface Implement Row Partition algorithm with DryadLINQ Performance study Stage Three Refinement experiments results Report and presentation

14 Backup slides

15  Input: C# and LINQ data objects  DryadLINQ distributed data objects.  DryadLINQ translates LINQ programs into distributed Dryad computations:  C# methods become code running on the vertices of a Dryad job.  Output: DryadLINQ distributed data objects .Net objects DryadLINQ Client machine (11) Distributed query plan.NET program Query Expr HPC Cluster Output Tables Results Input Tables Invoke Query Output DryadTable Dryad Execution.Net Objects JM ToTable foreach Vertex code Dryad Job Submission

16 Dryad Job Execution Flow

17 Performance on one Node

18 Performance on Multiple Node

19 Analysis for three algorithms

20 Performance for three algorithms Test done on 16 nodes of Tempest, using one core per node.

21 Performance for Multithreaded MM Test done on one node of Tempest, 24 cores


Download ppt "MATRIX MULTIPLY WITH DRYAD B649 Course Project Introduction."

Similar presentations


Ads by Google