swapping evaluation a memory scalable solution for answer
play

Swapping Evaluation: A Memory-Scalable Solution for Answer-On-Demand - PowerPoint PPT Presentation

Apr 10, 2023 •368 likes •775 views

Swapping Evaluation: A Memory-Scalable Solution for Answer-On-Demand Tabling an 1 Manuel Carro 1 David S. Warren 2 Pablo Chico de Guzm 1 School of Computer Science, Technical University of Madrid, Spain 2 State University of New York at Stony

  1. Swapping Evaluation: A Memory-Scalable Solution for Answer-On-Demand Tabling an 1 Manuel Carro 1 David S. Warren 2 Pablo Chico de Guzm´ 1 School of Computer Science, Technical University of Madrid, Spain 2 State University of New York at Stony Brook, USA ICLP 2010 — Edinburgh – July 16, 2010 Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 1 / 15

  2. Introduction Tabling Basics Tabling : resolution strategy (alternative to SLD) aimed at improving both efficiency and declarativeness in Prolog. ◮ ≈ memoing: answers for already executed calls retrieved from table. ◮ Infinite recursions automatically avoided. ?− p(A). Example code :- table p/1. q, p(B), A = 2. A = 1. p(A) :- p(B), A = 2. q, p(B), A = 2. Suspension p(1), A = 2. p(A) :- A = 1. A = 2. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 2 / 15

  3. Introduction External and Internal Consumers Internal: appears under its generator execution tree. External: appears out of the scope of its generator execution tree. Example code a b a ?- a, b, a. :- table a/0. a true EXTERNAL a :- a. a. b. INTERNAL Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 3 / 15

  4. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. comp1 t(X) :- computation2, X = 2. t(1) p :- large1. p :- large2. ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  5. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. comp1 t(X) :- computation2, X = 2. t(1) p :- large1. p :- large2. fail ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  6. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. comp2 t(X) :- computation2, X = 2. t(2) p :- large1. p :- large2. ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  7. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. comp2 t(X) :- computation2, X = 2. t(2) p :- large1. p :- large2. fail ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  8. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. t(X) :- computation1, X = 1. t(X) :- computation2, X = 2. p :- large1. p :- large2. ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  9. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. large1 t(X) :- computation2, t(1) X = 2. p :- large1. p :- large2. ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  10. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. large1 t(X) :- computation2, t(1) t(1) X = 2. p :- large1. p :- large2. ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  11. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. large1 t(X) :- computation2, t(2) t(1) X = 2. p :- large1. p :- large2. ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  12. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. large2 t(X) :- computation2, t(1) t(2) t(1) X = 2. p :- large1. p :- large2. ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  13. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. t(X) :- computation2, largeN X = 2. t(1) t(2) t(1) p :- large1. p :- large2. ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  14. Scheduling Strategy Tabling: Scheduling Strategy Local scheduling Computes all the answers before returning any of them. Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. t(X) :- computation2, large1 X = 2. t(2) t(1) t(2) p :- large1. p :- large2. ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 4 / 15

  15. Scheduling Strategy Tabling: Scheduling Strategy Batched scheduling Returns each of the answer as soon as they are computed. ◮ Cut-Once operators can prune when the desired answer is found. ◮ Bad memory behavior (live environments & external suspensions). Example code ?- t(X), p, t(Y), fail. p t(X) t(Y) fail :- table t/1. t(X) :- computation1, X = 1. comp1 t(X) :- computation2, X = 2. t(1) p :- large1. p :- large2. ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 5 / 15

  16. Scheduling Strategy Tabling: Scheduling Strategy Batched scheduling Returns each of the answer as soon as they are computed. ◮ Cut-Once operators can prune when the desired answer is found. ◮ Bad memory behavior (live environments & external suspensions). Example code ?- t(X), p, t(Y), fail. p t(X) t(Y) fail :- table t/1. t(X) :- computation1, X = 1. comp1 t(X) :- computation2, X = 2. t(1) p :- large1. p :- large2. large1 ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 5 / 15

  17. Scheduling Strategy Tabling: Scheduling Strategy Batched scheduling Returns each of the answer as soon as they are computed. ◮ Cut-Once operators can prune when the desired answer is found. ◮ Bad memory behavior (live environments & external suspensions). Example code ?- t(X), p, t(Y), fail. p t(X) t(Y) fail :- table t/1. t(X) :- computation1, t(1) X = 1. comp1 t(X) :- computation2, X = 2. t(1) p :- large1. p :- large2. large1 ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 5 / 15

  18. Scheduling Strategy Tabling: Scheduling Strategy Batched scheduling Returns each of the answer as soon as they are computed. ◮ Cut-Once operators can prune when the desired answer is found. ◮ Bad memory behavior (live environments & external suspensions). Example code ?- t(X), p, t(Y), fail. p t(X) t(Y) fail :- table t/1. t(X) :- computation1, t(1) X = 1. comp1 t(X) :- computation2, Susp. X = 2. t(1) p :- large1. p :- large2. large1 ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 5 / 15

  19. Scheduling Strategy Tabling: Scheduling Strategy Batched scheduling Returns each of the answer as soon as they are computed. ◮ Cut-Once operators can prune when the desired answer is found. ◮ Bad memory behavior (live environments & external suspensions). Example code ?- t(X), p, t(Y), fail. p t(X) t(Y) fail :- table t/1. t(X) :- computation1, t(1) X = 1. comp1 t(X) :- computation2, Susp. X = 2. t(1) p :- large1. p :- large2. large1 large2 ... p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 5 / 15

  20. Scheduling Strategy Tabling: Scheduling Strategy Batched scheduling Returns each of the answer as soon as they are computed. ◮ Cut-Once operators can prune when the desired answer is found. ◮ Bad memory behavior (live environments & external suspensions). Example code ?- t(X), p, t(Y), fail. :- table t/1. p t(X) t(Y) fail t(X) :- computation1, X = 1. t(1) comp1 t(X) :- computation2, X = 2. Susp. p :- large1. t(1) p :- large2. ... ... large1 large2 largeN p :- largeN. Pablo Chico de Guzm´ an et al (UPM) Swapping Evaluation. . . ICLP 2010, Edinburgh 5 / 15

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Swapping ! Active processes use more physical memory than system has Address Binding Swap out

Swapping ! Active processes use more physical memory than system has Address Binding Swap out

Swapping ! Active processes use more physical memory than system has Address Binding Swap out can be fixed P1 or relocatable Operating Systems I at runtime P2 Backing Store (Swap Space) OS Virtual Memory Main Memory Swapping

539 views • 8 slides
Chapter 8: Main Memory Chapter 8: Memory Management Background Swapping

Chapter 8: Main Memory Chapter 8: Memory Management Background Swapping

Chapter 8: Main Memory Chapter 8: Memory Management Background Swapping Contiguous Memory Allocation Segmentation Paging Structure of the Page Table Example: The Intel 32 and 64-bit Architectures

936 views • 72 slides
Operating System Principles: Memory Management Swapping, Paging, and Virtual Memory CS 111

Operating System Principles: Memory Management Swapping, Paging, and Virtual Memory CS 111

Operating System Principles: Memory Management Swapping, Paging, and Virtual Memory CS 111 Operating Systems Peter Reiher Lecture 6 CS 111 Page 1 Fall 2016 Outline Swapping Paging Virtual memory Lecture 6 CS 111 Page 2

744 views • 57 slides
Main Memory Management organizing memory hardware Swapping (which are pertinent to

Main Memory Management organizing memory hardware Swapping (which are pertinent to

Objectives/Outline Chapter 8: Objectives Outline Describe various ways of Background Main Memory Management organizing memory hardware Swapping (which are pertinent to Contiguous Allocation various memory managing

732 views • 16 slides
Using Tensor Swapping and NVLink to Overcome GPU Memory Limits with TensorFlow Sam Matzek Deep

Using Tensor Swapping and NVLink to Overcome GPU Memory Limits with TensorFlow Sam Matzek Deep

Using Tensor Swapping and NVLink to Overcome GPU Memory Limits with TensorFlow Sam Matzek Deep learning is memory constrained GPUs have limited memory Neural networks are growing deeper and wider Amount and size of data to process is

397 views • 23 slides
Memory Management Logical vs. physical address space Fragmentation Paging

Memory Management Logical vs. physical address space Fragmentation Paging

CPSC 410/611 : Operating Systems Memory Management Logical vs. physical address space Fragmentation Paging Segmentation An expensive way to run multiple processes: Swapping OS swap_out swap_in start swapping store memory

228 views • 12 slides
Virtual Memory and Paging 6A. Introduction to Swapping and Paging 6B. Paging MMUs and Demand

Virtual Memory and Paging 6A. Introduction to Swapping and Paging 6B. Paging MMUs and Demand

4/12/2016 Virtual Memory and Paging 6A. Introduction to Swapping and Paging 6B. Paging MMUs and Demand Paging Operating Systems Principles 6C. Replacement Algorithms Virtual Memory and Paging 6D. Thrashing and Working Sets 6E. Other

205 views • 8 slides
Memory Virtualization: Swapping and Demand Paging Policies 1 University of New Mexico Beyond

Memory Virtualization: Swapping and Demand Paging Policies 1 University of New Mexico Beyond

University of New Mexico Memory Virtualization: Swapping and Demand Paging Policies 1 University of New Mexico Beyond Physical Memory: Policies Memory pressure forces the OS to start paging out pages to make room for actively-used pages.

510 views • 24 slides
Memory Virtualization: Swapping and Demand Paging Mechanisms Prof. Patrick G. Bridges 1

Memory Virtualization: Swapping and Demand Paging Mechanisms Prof. Patrick G. Bridges 1

University of New Mexico Memory Virtualization: Swapping and Demand Paging Mechanisms Prof. Patrick G. Bridges 1 University of New Mexico Beyond Physical Memory: Mechanisms Require an additional level in the memory hierarchy. OS need

245 views • 12 slides
Cache Coherence in Scalable Machines Scalable Cache Coherent Systems Scalable, distributed

Cache Coherence in Scalable Machines Scalable Cache Coherent Systems Scalable, distributed

Cache Coherence in Scalable Machines Scalable Cache Coherent Systems Scalable, distributed memory plus coherent replication Scalable distributed memory machines P-C-M nodes connected by network communication assist interprets

1.13k views • 87 slides
Chapter 9: Memory Management Background Swapping Contiguous Allocation Paging

Chapter 9: Memory Management Background Swapping Contiguous Allocation Paging

Chapter 9: Memory Management Background Swapping Contiguous Allocation Paging Segmentation Segmentation with Paging Operating System Concepts Silberschatz, Galvin and Gagne 2002 9.1 Background Program must be

718 views • 27 slides
Chapter 9: Memory Management Background Swapping Contiguous Allocation Paging

Chapter 9: Memory Management Background Swapping Contiguous Allocation Paging

Chapter 9: Memory Management Background Swapping Contiguous Allocation Paging Segmentation Segmentation with Paging Operating System Concepts 9.1 Silberschatz, Galvin and Gagne 2002 Background Program must be

1.09k views • 54 slides
Module 8: Memory Management Background Logical versus Physical Address Space Swapping

Module 8: Memory Management Background Logical versus Physical Address Space Swapping

Module 8: Memory Management Background Logical versus Physical Address Space Swapping Contiguous Allocation Paging Segmentation Segmentation with Paging Operating System Concepts 8.1 Silberschatz and Galvin 1999

913 views • 41 slides
Module 8: Memory Management Background Logical versus Physical Address Space Swapping

Module 8: Memory Management Background Logical versus Physical Address Space Swapping

' $ Module 8: Memory Management Background Logical versus Physical Address Space Swapping Contiguous Allocation Paging Segmentation Segmentation with Paging & % Silberschatz and Galvin c Operating System

865 views • 41 slides
Memory Management One of the most important OS jobs. Review memory hierarchy sizes, speed,

Memory Management One of the most important OS jobs. Review memory hierarchy sizes, speed,

Memory Management One of the most important OS jobs. Review memory hierarchy sizes, speed, cost Growing memories -> growing programs need for swapping and paging The simplest way to use memory is to have one program in memory sharing

576 views • 16 slides
Memory Management What to do when coalescing fails 5H. Memory Compaction garbage collection

Memory Management What to do when coalescing fails 5H. Memory Compaction garbage collection

4/15/2018 Memory Management What to do when coalescing fails 5H. Memory Compaction garbage collection is just another way to free 6A. Swapping to secondary storage doesnt greatly help or hurt fragmentation 5E. Dynamic Relocation

346 views • 10 slides
outdoor learning experiences with children Hanin Hussain, PhD Early Childhood and Special Needs

outdoor learning experiences with children Hanin Hussain, PhD Early Childhood and Special Needs

Designing and enacting outdoor learning experiences with children Hanin Hussain, PhD Early Childhood and Special Needs Education Our itinerary for today Introducing our imaginary practice-based inquiry Phase 1: Inquiring into our starting

641 views • 53 slides
Optimization++ Complexities and strategies of optimization Instruction Scheduling

Optimization++ Complexities and strategies of optimization Instruction Scheduling

Optimization++ Complexities and strategies of optimization Instruction Scheduling Register Allocation Optimization Recap 1. Intermediate language (IL) module - better separation of front-end and back-end modules - permit

583 views • 16 slides
High level OCaml optimisations Pierre Chambart, OCamlPro OCaml 2013, 23 September 2013 OCaml is

High level OCaml optimisations Pierre Chambart, OCamlPro OCaml 2013, 23 September 2013 OCaml is

High level OCaml optimisations Pierre Chambart, OCamlPro OCaml 2013, 23 September 2013 OCaml is fast Not an optimising compiler, but: Predictable performances Good generated code What can we do to get faster ? Small modification on high

285 views • 25 slides
"Physics of the Deuteron" Group: Blaine Norum, Professor Richard Lindgren, Research

"Physics of the Deuteron" Group: Blaine Norum, Professor Richard Lindgren, Research

"Physics of the Deuteron" Group: Blaine Norum, Professor Richard Lindgren, Research Professor Kebin Wang, Research Scientist Sam Hoblit, Research Scientist Khamit Ardashev, Research Associate Search underway, Research Associate

692 views • 16 slides
15-780 Graduate Artificial Intelligence: Convolutional and recurrent networks J. Zico Kolter

15-780 Graduate Artificial Intelligence: Convolutional and recurrent networks J. Zico Kolter

15-780 Graduate Artificial Intelligence: Convolutional and recurrent networks J. Zico Kolter (this lecture) and Ariel Procaccia Carnegie Mellon University Spring 2017 1 Outline Convolutional neural networks Applications of convolutional

1.32k views • 32 slides
Multiplicity Computing Engineering Software for Reliability, Performance, and Security Alexander

Multiplicity Computing Engineering Software for Reliability, Performance, and Security Alexander

Multiplicity Computing Engineering Software for Reliability, Performance, and Security Alexander Wolf Department of Computing Imperial College London in collaboration with Cristian Cadar, Paolo Costa, and Peter Pietzuch An inspiring example

243 views • 13 slides
VisTrails: Visualization meets Data Management Erik Anderson, Steven Callahan, Juliana Freire,

VisTrails: Visualization meets Data Management Erik Anderson, Steven Callahan, Juliana Freire,

VisTrails: Visualization meets Data Management Erik Anderson, Steven Callahan, Juliana Freire, Emanuele Santos, Cludio Silva, Carlos Scheidegger, Nathan Smith, Huy Vo Visualization Large Amounts of Data Scientific Exploration

227 views • 6 slides
1 Topics (cont) Loop Transformations VI. Parallelism and Locality (*) Original code

1 Topics (cont) Loop Transformations VI. Parallelism and Locality (*) Original code

Final Review Questions to think about Today Possible big picture questions Overview of what we have learned so far For this class, which of the five criteria for evaluating optimizations have SSA review we used and how? (safety,

450 views • 4 slides

More recommend