3.2 Cache Oblivious Algorithms

Size: px
Start display at page:

Download "3.2 Cache Oblivious Algorithms"

Transcription

1 3.2 Cache Oblivious Algorithms

2 Cache-Oblivious Algorithms by Matteo Frigo, Charles E. Leiserson, Harald Prokop, and Sridhar Ramachandran. In the 40th Annual Symposium on Foundations of Computer Science, FOCS '99, October, 1999, New York, NY, USA. 2

3 Outline Cache complexity Cache aware algorithms Cache oblivious algorithms Matrix multiplication Matrix transposition FFT Conclusion 3

4 Assumption Only two levels of memory hierarchies: An ideal cache Fully associative Optimal replacement strategy Tall cache A very large memory 4

5 An Ideal Cache Model An ideal cache model (Z,L) Z: Total words in the cache L: Words in one cache line 5

6 Cache Complexity An algorithm with input size n is measured by: Work complexity W(n) Cache complexity: the number of cache misses it incurs. Q(n; Z, L) 6

7 Outline Cache complexity Cache aware algorithms Cache oblivious algorithms Matrix multiplication Matrix transposition FFT Conclusion 7

8 Cache Aware Algorithms Contain parameters to minimize the cache complexity for a particular cache size (Z) and line length (L). Need to adjust parameters when running on different platforms. 8

9 Example: A blocked matrix multiplication algorithm s s A11 n A s is a tuning parameter to make the algorithm run fast 9

10 Example (2) Cache complexity The three s x s sub matrices should fit into the cache so 2 2 they occupy max( s, s / L) = Θ( s + s / L) cache lines Optimal performance is obtained when Z/L cache misses needed to bring 3 sub matrices into cache n 2 /L cache misses needed to read n 2 elements 2 3 It is Θ(1 + n / L + ( n / s) ( Z / L)) = Θ(1 + n 2 / L + n 3 / L Z ) s = Θ( Z ) 10

11 Outline Cache complexity Cache aware algorithms Cache oblivious algorithms Matrix multiplication Matrix transposition and FFT Conclusion 11

12 Cache Oblivious Algorithms Have no parameters about hardware, such as cache size (Z), cache-line length (L). No tuning needed, platform independent. The following algorithms introduced are proved to have the optimal cache complexity. 12

13 Matrix Multiplication Partition matrix A and B by half in the largest dimension. A: n x m, B: m x p n max (m, p) m max (n, p) p max (n, m) Proceed recursively until reach the base case - one element. 13

14 Matrix Multiplication (2) Assume Sizes of A, B are nx4n, 4nxn B 11 ( A ) 11 A12 B12 A1*B1 A*B + A2*B2 + + B 1 ( A ) 1 A2 B2 21 ( A ) 21 A22 B22 A11*B11 A12*B12 A21*B21 A22*B22 B 14

15 Matrix Multiplication (3) Intuitively, once a sub problem fits into the cache, its smaller sub problems can be solved in cache with no further misses. 15

16 Matrix Multiplication (4) Cache complexity Can achieve the same as the cache complexity of Block-MULT algorithm (cache aware) For a square matrix, the optimal cache complexity is achieved. 16

17 Outline Cache complexity Cache aware algorithms Cache oblivious algorithms Matrix multiplication Matrix transposition FFT Conclusion 17

18 Matrix Transposition A A T for i 1 to m m x n B n x m for j 1 to n B( j, i ) = A( i, j ) If n is very large, the access of B in column will cause cache miss every time! (No spatial locality in B) 18

19 Matrix Transposition (2) Partition array A along the longer dimension and recursively execute the transpose function. A11 A21 A11 T A12 T A12 A22 A21 T A22 T 19

20 Matrix Transposition (3) Cache complexity It has the optimal cache complexity Q(m, n) = Θ(1+mn/L) 20

21 Fast Fourier Transform Y [ i ] = n 1 j = 0 X [ j ] ω ij n Use Cooley-Tukey algorithm Cooley-Tukey algorithms recursively re-express a DFT of a composite size n = n 1 n 2 as: Perform n 2 DFTs of size n 1. Multiply by complex roots of unity called twiddle factors. Perform n 1 DFTs of size n 2. 21

22 n 1 Yi [] = X[ j] w n2 1 n1 1 Yi [ 1+ in 2 1] = X[ jn j2] w w w j2= 0 j1= 0 j= 0 ij ij ij i j n n n n 1 n 2 22

23 Assume X is a row-major n 1 n 2 matrix Steps: Transpose X in place. Compute n 2 DFTs Multiply by twiddle factors Transpose X in place Compute n 1 DFTs Transpose X in-place 23

24 Fast Fourier Transform n1=4, n2=2 Transpose to select n2 DFT of size n1 Call FFT recursively with n1=2, n2=2 Reach the base case, return *twiddle factor Transpose to select n1 DFT of size n2 Transpose and return 24

25 Fast Fourier Transform Cache complexity Optimal for a Cooley-Tukey algorithm, when n is an exact power of 2 Q(n) = O(1+(n/L)(1+log z n) 25

26 Other Cache Oblivious Algorithms Funnelsort Distribution sort LU decomposition without pivots 26

27 Outline Cache complexity Cache aware algorithms Cache oblivious algorithms Matrix multiplication Matrix transposition FFT Conclusion 27

28 Questions How large is the range of practicality of cache-oblivious algorithms? What are the relative strengths of cacheoblivious and cache-aware algorithms? 28

29 Practicality of Cache-oblivious Algorithms Average time to transpose an NxN matrix, divided by N 2 29

30 Practicality of Cache-oblivious Algorithms (2) Average time taken to multiply two NxN matrices, divided by N 3 30

31 Question 2 Do cache-oblivious algorithms perform as well as cache-aware algorithms? FFTW library No answer yet. 31

32 References Cache-Oblivious Algorithms by Matteo Frigo, Charles E. Leiserson, Harald Prokop, and Sridhar Ramachandran. In the 40th Annual Symposium on Foundations of Computer Science, FOCS '99, October, 1999, New York, NY, USA. Cache-Oblivious Algorithms by Harald Prokop. Master's Thesis, MIT Department of Electrical Engineering and Computer Science. June Optimizing Matrix Multiplication with a Classifier Learning System by Xiaoming Li and María Jesus Garzarán. LCPC

Cache-Oblivious Algorithms

Cache-Oblivious Algorithms Cache-Oblivious Algorithms Paper Reading Group Matteo Frigo Charles E. Leiserson Harald Prokop Sridhar Ramachandran Presents: Maksym Planeta 03.09.2015 Table of Contents Introduction Cache-oblivious algorithms

More information

Cache-Oblivious Algorithms

Cache-Oblivious Algorithms Cache-Oblivious Algorithms Matteo Frigo, Charles Leiserson, Harald Prokop, Sridhar Ramchandran Slides Written and Presented by William Kuszmaul THE DISK ACCESS MODEL Three Parameters: B M P Block Size

More information

Cache Friendly Sparse Matrix Vector Multilication

Cache Friendly Sparse Matrix Vector Multilication Cache Friendly Sparse Matrix Vector Multilication Sardar Anisual Haque 1, Shahadat Hossain 2, Marc Moreno Maza 1 1 University of Western Ontario, London, Ontario (Canada) 2 Department of Computer Science,

More information

Memory Management Algorithms on Distributed Systems. Katie Becker and David Rodgers CS425 April 15, 2005

Memory Management Algorithms on Distributed Systems. Katie Becker and David Rodgers CS425 April 15, 2005 Memory Management Algorithms on Distributed Systems Katie Becker and David Rodgers CS425 April 15, 2005 Table of Contents 1. Introduction 2. Coarse Grained Memory 2.1. Bottlenecks 2.2. Simulations 2.3.

More information

Cache-Oblivious Algorithms EXTENDED ABSTRACT

Cache-Oblivious Algorithms EXTENDED ABSTRACT Cache-Oblivious Algorithms EXTENDED ABSTRACT Matteo Frigo Charles E. Leiserson Harald Prokop Sridhar Ramachandran MIT Laboratory for Computer Science, 545 Technology Square, Cambridge, MA 02139 fathena,cel,prokop,sridharg@supertech.lcs.mit.edu

More information

Cache-Efficient Algorithms

Cache-Efficient Algorithms 6.172 Performance Engineering of Software Systems LECTURE 8 Cache-Efficient Algorithms Charles E. Leiserson October 5, 2010 2010 Charles E. Leiserson 1 Ideal-Cache Model Recall: Two-level hierarchy. Cache

More information

Algorithms for dealing with massive data

Algorithms for dealing with massive data Computer Science Department Federal University of Rio Grande do Sul Porto Alegre, Brazil Outline of the talk Introduction Outline of the talk Algorithms models for dealing with massive datasets : Motivation,

More information

Cache-Adaptive Analysis

Cache-Adaptive Analysis Cache-Adaptive Analysis Michael A. Bender1 Erik Demaine4 Roozbeh Ebrahimi1 Jeremy T. Fineman3 Rob Johnson1 Andrea Lincoln4 Jayson Lynch4 Samuel McCauley1 1 3 4 Available Memory Can Fluctuate in Real Systems

More information

Cache-Oblivious Algorithms

Cache-Oblivious Algorithms Cache-Oblivious Algorithms by Harald Prokop Submitted to the Department of Electrical Engineering and Computer Science in partial fulfillment of the requirements for the degree of Master of Science at

More information

Algorithms and Data Structures: Efficient and Cache-Oblivious

Algorithms and Data Structures: Efficient and Cache-Oblivious 7 Ritika Angrish and Dr. Deepak Garg Algorithms and Data Structures: Efficient and Cache-Oblivious Ritika Angrish* and Dr. Deepak Garg Department of Computer Science and Engineering, Thapar University,

More information

Cache-Oblivious String Dictionaries

Cache-Oblivious String Dictionaries Cache-Oblivious String Dictionaries Gerth Stølting Brodal University of Aarhus Joint work with Rolf Fagerberg #"! Outline of Talk Cache-oblivious model Basic cache-oblivious techniques Cache-oblivious

More information

Funnel Heap - A Cache Oblivious Priority Queue

Funnel Heap - A Cache Oblivious Priority Queue Alcom-FT Technical Report Series ALCOMFT-TR-02-136 Funnel Heap - A Cache Oblivious Priority Queue Gerth Stølting Brodal, Rolf Fagerberg Abstract The cache oblivious model of computation is a two-level

More information

I/O Model. Cache-Oblivious Algorithms : Algorithms in the Real World. Advantages of Cache-Oblivious Algorithms 4/9/13

I/O Model. Cache-Oblivious Algorithms : Algorithms in the Real World. Advantages of Cache-Oblivious Algorithms 4/9/13 I/O Model 15-853: Algorithms in the Real World Locality II: Cache-oblivious algorithms Matrix multiplication Distribution sort Static searching Abstracts a single level of the memory hierarchy Fast memory

More information

Report Seminar Algorithm Engineering

Report Seminar Algorithm Engineering Report Seminar Algorithm Engineering G. S. Brodal, R. Fagerberg, K. Vinther: Engineering a Cache-Oblivious Sorting Algorithm Iftikhar Ahmad Chair of Algorithm and Complexity Department of Computer Science

More information

Cache Oblivious Matrix Transposition: Simulation and Experiment

Cache Oblivious Matrix Transposition: Simulation and Experiment Cache Oblivious Matrix Transposition: Simulation and Experiment Dimitrios Tsifakis, Alistair P. Rendell * and Peter E. Strazdins Department of Computer Science Australian National University Canberra ACT0200,

More information

Cache Oblivious Matrix Transpositions using Sequential Processing

Cache Oblivious Matrix Transpositions using Sequential Processing IOSR Journal of Engineering (IOSRJEN) e-issn: 225-321, p-issn: 2278-8719 Vol. 3, Issue 11 (November. 213), V4 PP 5-55 Cache Oblivious Matrix s using Sequential Processing korde P.S., and Khanale P.B 1

More information

Experimenting with the MetaFork Framework Targeting Multicores

Experimenting with the MetaFork Framework Targeting Multicores Experimenting with the MetaFork Framework Targeting Multicores Xiaohui Chen, Marc Moreno Maza & Sushek Shekar University of Western Ontario 26 January 2014 1 Introduction The work reported in this report

More information

Cache Oblivious Algorithms

Cache Oblivious Algorithms Cache Oblivious Algorithms Volker Strumpen IBM Research Austin, TX September 4, 2007 Iterative matrix transposition #define N 1000 double A[N][N], B[N][N]; void iter(void) { int i, j; for (i = 0; i < N;

More information

Cache Memories, Cache Complexity

Cache Memories, Cache Complexity Cache Memories, Cache Complexity Marc Moreno Maza University of Western Ontario, London (Canada) Applications of Computer Algebra Session on High-Performance Computer Algebra Jerusalem College of Technology,

More information

Cache-Oblivious Traversals of an Array s Pairs

Cache-Oblivious Traversals of an Array s Pairs Cache-Oblivious Traversals of an Array s Pairs Tobias Johnson May 7, 2007 Abstract Cache-obliviousness is a concept first introduced by Frigo et al. in [1]. We follow their model and develop a cache-oblivious

More information

Adaptive Matrix Transpose Algorithms for Distributed Multicore Processors

Adaptive Matrix Transpose Algorithms for Distributed Multicore Processors Adaptive Matrix Transpose Algorithms for Distributed Multicore ors John C. Bowman and Malcolm Roberts Abstract An adaptive parallel matrix transpose algorithm optimized for distributed multicore architectures

More information

Cache Oblivious Matrix Transposition: Simulation and Experiment

Cache Oblivious Matrix Transposition: Simulation and Experiment Cache Oblivious Matrix Transposition: Simulation and Experiment Dimitrios Tsifakis, Alistair P. Rendell, and Peter E. Strazdins Department of Computer Science, Australian National University Canberra ACT0200,

More information

Cache Memories, Cache Complexity

Cache Memories, Cache Complexity Cache Memories, Cache Complexity Marc Moreno Maza University of Western Ontario, London, Ontario (Canada) CS3101 and CS4402-9535 Plan Hierarchical memories and their impact on our programs Cache Analysis

More information

Network-oblivious algorithms. Gianfranco Bilardi, Andrea Pietracaprina, Geppino Pucci and Francesco Silvestri

Network-oblivious algorithms. Gianfranco Bilardi, Andrea Pietracaprina, Geppino Pucci and Francesco Silvestri Network-oblivious algorithms Gianfranco Bilardi, Andrea Pietracaprina, Geppino Pucci and Francesco Silvestri Overview Motivation Framework for network-oblivious algorithms Case studies: Network-oblivious

More information

Cache-efficient string sorting for Burrows-Wheeler Transform. Advait D. Karande Sriram Saroop

Cache-efficient string sorting for Burrows-Wheeler Transform. Advait D. Karande Sriram Saroop Cache-efficient string sorting for Burrows-Wheeler Transform Advait D. Karande Sriram Saroop What is Burrows-Wheeler Transform? A pre-processing step for data compression Involves sorting of all rotations

More information

Cache Memories. University of Western Ontario, London, Ontario (Canada) Marc Moreno Maza. CS2101 October 2012

Cache Memories. University of Western Ontario, London, Ontario (Canada) Marc Moreno Maza. CS2101 October 2012 Cache Memories Marc Moreno Maza University of Western Ontario, London, Ontario (Canada) CS2101 October 2012 Plan 1 Hierarchical memories and their impact on our 2 Cache Analysis in Practice Plan 1 Hierarchical

More information

Lecture 24 November 24, 2015

Lecture 24 November 24, 2015 CS 229r: Algorithms for Big Data Fall 2015 Prof. Jelani Nelson Lecture 24 November 24, 2015 Scribes: Zhengyu Wang 1 Cache-oblivious Model Last time we talked about disk access model (as known as DAM, or

More information

How to Write Fast Numerical Code Spring 2012 Lecture 20. Instructor: Markus Püschel TAs: Georg Ofenbeck & Daniele Spampinato

How to Write Fast Numerical Code Spring 2012 Lecture 20. Instructor: Markus Püschel TAs: Georg Ofenbeck & Daniele Spampinato How to Write Fast Numerical Code Spring 2012 Lecture 20 Instructor: Markus Püschel TAs: Georg Ofenbeck & Daniele Spampinato Planning Today Lecture Project meetings Project presentations 10 minutes each

More information

Adaptive Matrix Transpose Algorithms for Distributed Multicore Processors

Adaptive Matrix Transpose Algorithms for Distributed Multicore Processors Adaptive Matrix Transpose Algorithms for Distributed Multicore ors John C. Bowman and Malcolm Roberts Abstract An adaptive parallel matrix transpose algorithm optimized for distributed multicore architectures

More information

Cache Oblivious Stencil Computations

Cache Oblivious Stencil Computations Cache Oblivious Stencil Computations Matteo Frigo and Volker Strumpen IBM Austin Research Laboratory 11501 Burnet Road, Austin, TX 78758 May 25, 2005 Abstract We present a cache oblivious algorithm for

More information

How to Write Fast Numerical Code Spring 2011 Lecture 22. Instructor: Markus Püschel TA: Georg Ofenbeck

How to Write Fast Numerical Code Spring 2011 Lecture 22. Instructor: Markus Püschel TA: Georg Ofenbeck How to Write Fast Numerical Code Spring 2011 Lecture 22 Instructor: Markus Püschel TA: Georg Ofenbeck Schedule Today Lecture Project presentations 10 minutes each random order random speaker 10 Final code

More information

Outline of the talk 1 Problem definition 2 Computational Models 3 Technical results 4 Conclusions University of Padova Bertinoro, February 17-18th 200

Outline of the talk 1 Problem definition 2 Computational Models 3 Technical results 4 Conclusions University of Padova Bertinoro, February 17-18th 200 Cache-Oblivious Simulation of Parallel Programs Andrea Pietracaprina Geppino Pucci Francesco Silvestri Bertinoro, February 17-18th 2006 University of Padova Bertinoro, February 17-18th 2006 1/ 19 Outline

More information

Cache-Oblivious Algorithms A Unified Approach to Hierarchical Memory Algorithms

Cache-Oblivious Algorithms A Unified Approach to Hierarchical Memory Algorithms Cache-Oblivious Algorithms A Unified Approach to Hierarchical Memory Algorithms Aarhus University Cache-Oblivious Current Trends Algorithms in Algorithms, - A Unified Complexity Approach to Theory, Hierarchical

More information

Cache-Oblivious Algorithms and Data Structures

Cache-Oblivious Algorithms and Data Structures Cache-Oblivious Algorithms and Data Structures Erik D. Demaine MIT Laboratory for Computer Science, 200 Technology Square, Cambridge, MA 02139, USA, edemaine@mit.edu Abstract. A recent direction in the

More information

Effect of memory latency

Effect of memory latency CACHE AWARENESS Effect of memory latency Consider a processor operating at 1 GHz (1 ns clock) connected to a DRAM with a latency of 100 ns. Assume that the processor has two ALU units and it is capable

More information

Lecture 19 Apr 25, 2007

Lecture 19 Apr 25, 2007 6.851: Advanced Data Structures Spring 2007 Prof. Erik Demaine Lecture 19 Apr 25, 2007 Scribe: Aditya Rathnam 1 Overview Previously we worked in the RA or cell probe models, in which the cost of an algorithm

More information

The History of I/O Models Erik Demaine

The History of I/O Models Erik Demaine The History of I/O Models Erik Demaine MASSACHUSETTS INSTITUTE OF TECHNOLOGY Memory Hierarchies in Practice CPU 750 ps Registers 100B Level 1 Cache 100KB Level 2 Cache 1MB 10GB 14 ns Main Memory 1EB-1ZB

More information

Introducing a Cache-Oblivious Blocking Approach for the Lattice Boltzmann Method

Introducing a Cache-Oblivious Blocking Approach for the Lattice Boltzmann Method Introducing a Cache-Oblivious Blocking Approach for the Lattice Boltzmann Method G. Wellein, T. Zeiser, G. Hager HPC Services Regional Computing Center A. Nitsure, K. Iglberger, U. Rüde Chair for System

More information

Parallel FFT Program Optimizations on Heterogeneous Computers

Parallel FFT Program Optimizations on Heterogeneous Computers Parallel FFT Program Optimizations on Heterogeneous Computers Shuo Chen, Xiaoming Li Department of Electrical and Computer Engineering University of Delaware, Newark, DE 19716 Outline Part I: A Hybrid

More information

How to Write Fast Numerical Code

How to Write Fast Numerical Code How to Write Fast Numerical Code Lecture: Optimizing FFT, FFTW Instructor: Markus Püschel TA: Georg Ofenbeck & Daniele Spampinato Rest of Semester Today Lecture Project meetings Project presentations 10

More information

Network-Oblivious Algorithms. Gianfranco Bilardi, Andrea Pietracaprina, Geppino Pucci, Michele Scquizzato and Francesco Silvestri

Network-Oblivious Algorithms. Gianfranco Bilardi, Andrea Pietracaprina, Geppino Pucci, Michele Scquizzato and Francesco Silvestri Network-Oblivious Algorithms Gianfranco Bilardi, Andrea Pietracaprina, Geppino Pucci, Michele Scquizzato and Francesco Silvestri Overview Background Summary of results Framework for network-oblivious algorithms

More information

Dense Matrix Multiplication

Dense Matrix Multiplication Dense Matrix Multiplication Abhishek Somani, Debdeep Mukhopadhyay Mentor Graphics, IIT Kharagpur October 7, 2015 Abhishek, Debdeep (IIT Kgp) Matrix Mult. October 7, 2015 1 / 56 Overview 1 The Problem 2

More information

CS473 - Algorithms I

CS473 - Algorithms I CS473 - Algorithms I Lecture 4 The Divide-and-Conquer Design Paradigm View in slide-show mode 1 Reminder: Merge Sort Input array A sort this half sort this half Divide Conquer merge two sorted halves Combine

More information

Algorithms and Computation in Signal Processing

Algorithms and Computation in Signal Processing Algorithms and Computation in Signal Processing special topic course 18-799B spring 2005 14 th Lecture Feb. 24, 2005 Instructor: Markus Pueschel TA: Srinivas Chellappa Course Evaluation Email sent out

More information

Twiddle Factor Transformation for Pipelined FFT Processing

Twiddle Factor Transformation for Pipelined FFT Processing Twiddle Factor Transformation for Pipelined FFT Processing In-Cheol Park, WonHee Son, and Ji-Hoon Kim School of EECS, Korea Advanced Institute of Science and Technology, Daejeon, Korea icpark@ee.kaist.ac.kr,

More information

Input parameters System specifics, user options. Input parameters size, dim,... FFT Code Generator. Initialization Select fastest execution plan

Input parameters System specifics, user options. Input parameters size, dim,... FFT Code Generator. Initialization Select fastest execution plan Automatic Performance Tuning in the UHFFT Library Dragan Mirković 1 and S. Lennart Johnsson 1 Department of Computer Science University of Houston Houston, TX 7724 mirkovic@cs.uh.edu, johnsson@cs.uh.edu

More information

6.895 Final Project: Serial and Parallel execution of Funnel Sort

6.895 Final Project: Serial and Parallel execution of Funnel Sort 6.895 Final Project: Serial and Parallel execution of Funnel Sort Paul Youn December 17, 2003 Abstract The speed of a sorting algorithm is often measured based on the sheer number of calculations required

More information

How to Write Fast Numerical Code

How to Write Fast Numerical Code How to Write Fast Numerical Code Lecture: Cost analysis and performance Instructor: Markus Püschel TA: Gagandeep Singh, Daniele Spampinato & Alen Stojanov Technicalities Research project: Let us know (fastcode@lists.inf.ethz.ch)

More information

Energy Efficient Adaptive Beamforming on Sensor Networks

Energy Efficient Adaptive Beamforming on Sensor Networks Energy Efficient Adaptive Beamforming on Sensor Networks Viktor K. Prasanna Bhargava Gundala, Mitali Singh Dept. of EE-Systems University of Southern California email: prasanna@usc.edu http://ceng.usc.edu/~prasanna

More information

Multi-core Computing Lecture 2

Multi-core Computing Lecture 2 Multi-core Computing Lecture 2 MADALGO Summer School 2012 Algorithms for Modern Parallel and Distributed Models Phillip B. Gibbons Intel Labs Pittsburgh August 21, 2012 Multi-core Computing Lectures: Progress-to-date

More information

Report on Cache-Oblivious Priority Queue and Graph Algorithm Applications[1]

Report on Cache-Oblivious Priority Queue and Graph Algorithm Applications[1] Report on Cache-Oblivious Priority Queue and Graph Algorithm Applications[1] Marc André Tanner May 30, 2014 Abstract This report contains two main sections: In section 1 the cache-oblivious computational

More information

CS 140 : Numerical Examples on Shared Memory with Cilk++

CS 140 : Numerical Examples on Shared Memory with Cilk++ CS 140 : Numerical Examples on Shared Memory with Cilk++ Matrix-matrix multiplication Matrix-vector multiplication Hyperobjects Thanks to Charles E. Leiserson for some of these slides 1 Work and Span (Recap)

More information

Autotuning (1/2): Cache-oblivious algorithms

Autotuning (1/2): Cache-oblivious algorithms Autotuning (1/2): Cache-oblivious algorithms Prof. Richard Vuduc Georgia Institute of Technology CSE/CS 8803 PNA: Parallel Numerical Algorithms [L.17] Tuesday, March 4, 2008 1 Today s sources CS 267 (Demmel

More information

Multi-core Computing Lecture 1

Multi-core Computing Lecture 1 Hi-Spade Multi-core Computing Lecture 1 MADALGO Summer School 2012 Algorithms for Modern Parallel and Distributed Models Phillip B. Gibbons Intel Labs Pittsburgh August 20, 2012 Lecture 1 Outline Multi-cores:

More information

Lecture 8 13 March, 2012

Lecture 8 13 March, 2012 6.851: Advanced Data Structures Spring 2012 Prof. Erik Demaine Lecture 8 13 March, 2012 1 From Last Lectures... In the previous lecture, we discussed the External Memory and Cache Oblivious memory models.

More information

How to Write Fast Numerical Code

How to Write Fast Numerical Code How to Write Fast Numerical Code Lecture: Cost analysis and performance Instructor: Markus Püschel TA: Alen Stojanov, Georg Ofenbeck, Gagandeep Singh Technicalities Research project: Let us know (fastcode@lists.inf.ethz.ch)

More information

Scheduling FFT Computation on SMP and Multicore Systems Ayaz Ali, Lennart Johnsson & Jaspal Subhlok

Scheduling FFT Computation on SMP and Multicore Systems Ayaz Ali, Lennart Johnsson & Jaspal Subhlok Scheduling FFT Computation on SMP and Multicore Systems Ayaz Ali, Lennart Johnsson & Jaspal Subhlok Texas Learning and Computation Center Department of Computer Science University of Houston Outline Motivation

More information

Implementing FFTs in Practice

Implementing FFTs in Practice Connexions module: m16336 1 Implementing FFTs in Practice Steven G. Johnson Matteo Frigo This work is produced by The Connexions Project and licensed under the Creative Commons Attribution License Abstract

More information

CSE 638: Advanced Algorithms. Lectures 18 & 19 ( Cache-efficient Searching and Sorting )

CSE 638: Advanced Algorithms. Lectures 18 & 19 ( Cache-efficient Searching and Sorting ) CSE 638: Advanced Algorithms Lectures 18 & 19 ( Cache-efficient Searching and Sorting ) Rezaul A. Chowdhury Department of Computer Science SUNY Stony Brook Spring 2013 Searching ( Static B-Trees ) A Static

More information

Cache-oblivious comparison-based algorithms on multisets

Cache-oblivious comparison-based algorithms on multisets Cache-oblivious comparison-based algorithms on multisets Arash Farzan 1, Paolo Ferragina 2, Gianni Franceschini 2, and J. Ian unro 1 1 {afarzan, imunro}@uwaterloo.ca School of Computer Science, University

More information

CS3350B Computer Architecture

CS3350B Computer Architecture CS3350B Computer Architecture Winter 2015 Lecture 3.1: Memory Hierarchy: What and Why? Marc Moreno Maza www.csd.uwo.ca/courses/cs3350b [Adapted from lectures on Computer Organization and Design, Patterson

More information

Basic Communication Ops

Basic Communication Ops CS 575 Parallel Processing Lecture 5: Ch 4 (GGKK) Sanjay Rajopadhye Colorado State University Basic Communication Ops n PRAM, final thoughts n Quiz 3 n Collective Communication n Broadcast & Reduction

More information

Cache Efficient Simple Dynamic Programming

Cache Efficient Simple Dynamic Programming Cache Efficient Simple Dynamic Programming Cary Cherng Richard E. Ladner September 25, 2004 Abstract New cache-oblivious and cache-aware algorithms for simple dynamic programming based on Valiant s context-free

More information

Outline. CS38 Introduction to Algorithms. Fast Fourier Transform (FFT) Fast Fourier Transform (FFT) Fast Fourier Transform (FFT)

Outline. CS38 Introduction to Algorithms. Fast Fourier Transform (FFT) Fast Fourier Transform (FFT) Fast Fourier Transform (FFT) Outline CS8 Introduction to Algorithms Lecture 9 April 9, 0 Divide and Conquer design paradigm matrix multiplication Dynamic programming design paradigm Fibonacci numbers weighted interval scheduling knapsack

More information

Cache-Aware and Cache-Oblivious Adaptive Sorting

Cache-Aware and Cache-Oblivious Adaptive Sorting Cache-Aware and Cache-Oblivious Adaptive Sorting Gerth Stølting rodal 1,, Rolf Fagerberg 2,, and Gabriel Moruz 1 1 RICS, Department of Computer Science, University of Aarhus, IT Parken, Åbogade 34, DK-8200

More information

Lecture 9 March 15, 2012

Lecture 9 March 15, 2012 6.851: Advanced Data Structures Spring 2012 Prof. Erik Demaine Lecture 9 March 15, 2012 1 Overview This is the last lecture on memory hierarchies. Today s lecture is a crossover between cache-oblivious

More information

Cache-oblivious Programming

Cache-oblivious Programming Cache-oblivious Programming Story so far We have studied cache optimizations for array programs Main transformations: loop interchange, loop tiling Loop tiling converts matrix computations into block matrix

More information

Lecture April, 2010

Lecture April, 2010 6.851: Advanced Data Structures Spring 2010 Prof. Eri Demaine Lecture 20 22 April, 2010 1 Memory Hierarchies and Models of Them So far in class, we have wored with models of computation lie the word RAM

More information

CHAPTER 6 A SECURE FAST 2D-DISCRETE FRACTIONAL FOURIER TRANSFORM BASED MEDICAL IMAGE COMPRESSION USING SPIHT ALGORITHM WITH HUFFMAN ENCODER

CHAPTER 6 A SECURE FAST 2D-DISCRETE FRACTIONAL FOURIER TRANSFORM BASED MEDICAL IMAGE COMPRESSION USING SPIHT ALGORITHM WITH HUFFMAN ENCODER 115 CHAPTER 6 A SECURE FAST 2D-DISCRETE FRACTIONAL FOURIER TRANSFORM BASED MEDICAL IMAGE COMPRESSION USING SPIHT ALGORITHM WITH HUFFMAN ENCODER 6.1. INTRODUCTION Various transforms like DCT, DFT used to

More information

BRICS Research Activities Algorithms

BRICS Research Activities Algorithms BRICS Research Activities Algorithms Gerth Stølting Brodal BRICS Retreat, Sandbjerg, 21 23 October 2002 1 Outline of Talk The Algorithms Group Courses Algorithm Events Expertise within BRICS Examples Algorithms

More information

FFT ALGORITHMS FOR MULTIPLY-ADD ARCHITECTURES

FFT ALGORITHMS FOR MULTIPLY-ADD ARCHITECTURES FFT ALGORITHMS FOR MULTIPLY-ADD ARCHITECTURES FRANCHETTI Franz, (AUT), KALTENBERGER Florian, (AUT), UEBERHUBER Christoph W. (AUT) Abstract. FFTs are the single most important algorithms in science and

More information

Fast Algorithm for Matrix-Vector Multiply of Asymmetric Multilevel Block-Toeplitz Matrices

Fast Algorithm for Matrix-Vector Multiply of Asymmetric Multilevel Block-Toeplitz Matrices " Fast Algorithm for Matrix-Vector Multiply of Asymmetric Multilevel Block-Toeplitz Matrices B. E. Barrowes, F. L. Teixeira, and J. A. Kong Research Laboratory of Electronics, MIT, Cambridge, MA 02139-4307

More information

Computational Methods CMSC/AMSC/MAPL 460. Vectors, Matrices, Linear Systems, LU Decomposition, Ramani Duraiswami, Dept. of Computer Science

Computational Methods CMSC/AMSC/MAPL 460. Vectors, Matrices, Linear Systems, LU Decomposition, Ramani Duraiswami, Dept. of Computer Science Computational Methods CMSC/AMSC/MAPL 460 Vectors, Matrices, Linear Systems, LU Decomposition, Ramani Duraiswami, Dept. of Computer Science Zero elements of first column below 1 st row multiplying 1 st

More information

SDP Memo 048: Two Dimensional Sparse Fourier Transform Algorithms

SDP Memo 048: Two Dimensional Sparse Fourier Transform Algorithms SDP Memo 048: Two Dimensional Sparse Fourier Transform Algorithms Document Number......................................................... SDP Memo 048 Document Type.....................................................................

More information

Module 9 : Numerical Relaying II : DSP Perspective

Module 9 : Numerical Relaying II : DSP Perspective Module 9 : Numerical Relaying II : DSP Perspective Lecture 36 : Fast Fourier Transform Objectives In this lecture, We will introduce Fast Fourier Transform (FFT). We will show equivalence between FFT and

More information

System Demonstration of Spiral: Generator for High-Performance Linear Transform Libraries

System Demonstration of Spiral: Generator for High-Performance Linear Transform Libraries System Demonstration of Spiral: Generator for High-Performance Linear Transform Libraries Yevgen Voronenko, Franz Franchetti, Frédéric de Mesmay, and Markus Püschel Department of Electrical and Computer

More information

Image Processing. Application area chosen because it has very good parallelism and interesting output.

Image Processing. Application area chosen because it has very good parallelism and interesting output. Chapter 11 Slide 517 Image Processing Application area chosen because it has very good parallelism and interesting output. Low-level Image Processing Operates directly on stored image to improve/enhance

More information

Analysis of Multithreaded Algorithms

Analysis of Multithreaded Algorithms Analysis of Multithreaded Algorithms Marc Moreno Maza University of Western Ontario, London, Ontario (Canada) CS4402-9535 (Moreno Maza) Analysis of Multithreaded Algorithms CS4402-9535 1 / 27 Plan 1 Matrix

More information

An Efficient Architecture for Ultra Long FFTs in FPGAs and ASICs

An Efficient Architecture for Ultra Long FFTs in FPGAs and ASICs HPEC 2004 Abstract Submission Dillon Engineering, Inc. www.dilloneng.com An Efficient Architecture for Ultra Long FFTs in FPGAs and ASICs Tom Dillon Dillon Engineering, Inc. This presentation outlines

More information

FPGA Based Design and Simulation of 32- Point FFT Through Radix-2 DIT Algorith

FPGA Based Design and Simulation of 32- Point FFT Through Radix-2 DIT Algorith FPGA Based Design and Simulation of 32- Point FFT Through Radix-2 DIT Algorith Sudhanshu Mohan Khare M.Tech (perusing), Dept. of ECE Laxmi Naraian College of Technology, Bhopal, India M. Zahid Alam Associate

More information

x = 12 x = 12 1x = 16

x = 12 x = 12 1x = 16 2.2 - The Inverse of a Matrix We've seen how to add matrices, multiply them by scalars, subtract them, and multiply one matrix by another. The question naturally arises: Can we divide one matrix by another?

More information

External Memory. Philip Bille

External Memory. Philip Bille External Memory Philip Bille Outline Computationals models Modern computers (word) RAM I/O Cache-oblivious Shortest path in implicit grid graphs RAM algorithm I/O algorithms Cache-oblivious algorithm Computational

More information

38 Cache-Oblivious Data Structures

38 Cache-Oblivious Data Structures 38 Cache-Oblivious Data Structures Lars Arge Duke University Gerth Stølting Brodal University of Aarhus Rolf Fagerberg University of Southern Denmark 38.1 The Cache-Oblivious Model... 38-1 38.2 Fundamental

More information

Formal Loop Merging for Signal Transforms

Formal Loop Merging for Signal Transforms Formal Loop Merging for Signal Transforms Franz Franchetti Yevgen S. Voronenko Markus Püschel Department of Electrical & Computer Engineering Carnegie Mellon University This work was supported by NSF through

More information

Dynamic programming in faulty memory hierarchies (cache-obliviously)

Dynamic programming in faulty memory hierarchies (cache-obliviously) Dynamic programming in faulty memory hierarchies (cache-obliviously) S. Caminiti 1, I. Finocchi 1, E. G. Fusco 1, and F. Silvestri 2 1 Computer Science Department, Sapienza University of Rome 2 Department

More information

Module 5: Performance Issues in Shared Memory and Introduction to Coherence Lecture 9: Performance Issues in Shared Memory. The Lecture Contains:

Module 5: Performance Issues in Shared Memory and Introduction to Coherence Lecture 9: Performance Issues in Shared Memory. The Lecture Contains: The Lecture Contains: Data Access and Communication Data Access Artifactual Comm. Capacity Problem Temporal Locality Spatial Locality 2D to 4D Conversion Transfer Granularity Worse: False Sharing Contention

More information

FFT. There are many ways to decompose an FFT [Rabiner and Gold] The simplest ones are radix-2 Computation made up of radix-2 butterflies X = A + BW

FFT. There are many ways to decompose an FFT [Rabiner and Gold] The simplest ones are radix-2 Computation made up of radix-2 butterflies X = A + BW FFT There are many ways to decompose an FFT [Rabiner and Gold] The simplest ones are radix-2 Computation made up of radix-2 butterflies A X = A + BW B Y = A BW B. Baas 442 FFT Dataflow Diagram Dataflow

More information

The Design and Implementation of FFTW3

The Design and Implementation of FFTW3 The Design and Implementation of FFTW3 MATTEO FRIGO AND STEVEN G. JOHNSON Invited Paper FFTW is an implementation of the discrete Fourier transform (DFT) that adapts to the hardware in order to maximize

More information

arxiv: v1 [cs.ds] 7 May 2016 Abstract

arxiv: v1 [cs.ds] 7 May 2016 Abstract The I/O complexity of Strassen s matrix multiplication with recomputation Gianfranco Bilardi 1 and Lorenzo De Stefani 2 1 Department of Information Engineering, University of Padova, Via Gradenigo 6B/Padova,

More information

Algorithm Design and Analysis

Algorithm Design and Analysis Algorithm Design and Analysis LECTURE 13 Divide and Conquer Closest Pair of Points Convex Hull Strassen Matrix Mult. Adam Smith 9/24/2008 A. Smith; based on slides by E. Demaine, C. Leiserson, S. Raskhodnikova,

More information

Learning to Construct Fast Signal Processing Implementations

Learning to Construct Fast Signal Processing Implementations Journal of Machine Learning Research 3 (2002) 887-919 Submitted 12/01; Published 12/02 Learning to Construct Fast Signal Processing Implementations Bryan Singer Manuela Veloso Department of Computer Science

More information

Matrix Multiplication

Matrix Multiplication Matrix Multiplication CPS343 Parallel and High Performance Computing Spring 2013 CPS343 (Parallel and HPC) Matrix Multiplication Spring 2013 1 / 32 Outline 1 Matrix operations Importance Dense and sparse

More information

Introduction to Multithreaded Algorithms

Introduction to Multithreaded Algorithms Introduction to Multithreaded Algorithms CCOM5050: Design and Analysis of Algorithms Chapter VII Selected Topics T. H. Cormen, C. E. Leiserson, R. L. Rivest, C. Stein. Introduction to algorithms, 3 rd

More information

6. Fast Fourier Transform

6. Fast Fourier Transform x[] X[] x[] x[] x[6] X[] X[] X[3] x[] x[5] x[3] x[7] 3 X[] X[5] X[6] X[7] A Historical Perspective The Cooley and Tukey Fast Fourier Transform (FFT) algorithm is a turning point to the computation of DFT

More information

Cache-Oblivious and Data-Oblivious Sorting and Applications

Cache-Oblivious and Data-Oblivious Sorting and Applications Cache-Oblivious and Data-Oblivious Sorting and Applications T-H. Hubert Chan, Yue Guo, Wei-Kai Lin, and Elaine Shi Jan, 2018 External Memory Model Cache efficiency: # of blocks Time: # of words Memory

More information

High performance 2D Discrete Fourier Transform on Heterogeneous Platforms. Shrenik Lad, IIIT Hyderabad Advisor : Dr. Kishore Kothapalli

High performance 2D Discrete Fourier Transform on Heterogeneous Platforms. Shrenik Lad, IIIT Hyderabad Advisor : Dr. Kishore Kothapalli High performance 2D Discrete Fourier Transform on Heterogeneous Platforms Shrenik Lad, IIIT Hyderabad Advisor : Dr. Kishore Kothapalli Motivation Fourier Transform widely used in Physics, Astronomy, Engineering

More information

Assignment #6: Subspaces of R n, Bases, Dimension and Rank. Due date: Wednesday, October 26, 2016 (9:10am) Name: Section Number

Assignment #6: Subspaces of R n, Bases, Dimension and Rank. Due date: Wednesday, October 26, 2016 (9:10am) Name: Section Number Assignment #6: Subspaces of R n, Bases, Dimension and Rank Due date: Wednesday, October 26, 206 (9:0am) Name: Section Number Assignment #6: Subspaces of R n, Bases, Dimension and Rank Due date: Wednesday,

More information

Cache efficient simple dynamic programming

Cache efficient simple dynamic programming Cache efficient simple dynamic programming Cary Cherng, Richard E. Ladner To cite this version: Cary Cherng, Richard E. Ladner. Cache efficient simple dynamic programming. Conrado Martínez. 2005 International

More information

Communication Efficient Gaussian Elimination with Partial Pivoting using a Shape Morphing Data Layout

Communication Efficient Gaussian Elimination with Partial Pivoting using a Shape Morphing Data Layout Communication Efficient Gaussian Elimination with Partial Pivoting using a Shape Morphing Data Layout Grey Ballard James Demmel Benjamin Lipshitz Oded Schwartz Sivan Toledo Electrical Engineering and Computer

More information

Plan. 1 Parallelism Complexity Measures. 2 cilk for Loops. 3 Scheduling Theory and Implementation. 4 Measuring Parallelism in Practice

Plan. 1 Parallelism Complexity Measures. 2 cilk for Loops. 3 Scheduling Theory and Implementation. 4 Measuring Parallelism in Practice lan Multithreaded arallelism and erformance Measures Marc Moreno Maza University of Western Ontario, London, Ontario (Canada) CS 4435 - CS 9624 1 2 cilk for Loops 3 4 Measuring arallelism in ractice 5

More information