DB system design for new hardware and sciences

9
DB system design for new hardware and sciences Anastasia Ailamaki École Polytechnique Fédérale de Lausanne and Carnegie Mellon University

description

DB system design for new hardware and sciences. Anastasia Ailamaki École Polytechnique Fédérale de Lausanne and Carnegie Mellon University. introspective. sharing. Uni-processor. Multi-core. Multi-processor. Cluster. Exploit max parallelism and sharing simultaneously. parallelism. - PowerPoint PPT Presentation

Transcript of DB system design for new hardware and sciences

Page 1: DB system design for new hardware and sciences

DB system design for new hardware and sciences

Anastasia Ailamaki

École Polytechnique Fédérale de Lausanneand

Carnegie Mellon University

Page 2: DB system design for new hardware and sciences

introspective

Page 3: DB system design for new hardware and sciences

parallelism

shar

ing

Uni-processor Multi-core

Multi-processor Cluster

Exploit max parallelism and sharing simultaneously

Page 4: DB system design for new hardware and sciences

Moore’s Law = cores

Performance burden shifts to software

0

50

100

150

200

1985 1990 1995 2000 2005 2010

Year of Introduction

Nu

mb

er o

f H

W T

hre

ads

UltraSparc T2

Power4

80486

CRS-1 (Tensilica)

Page 5: DB system design for new hardware and sciences

Multi-core challenges for DBMS

• CMP-aware parallelism in OLTP – Efficient synchronization– Highly concurrent algorithms

• CMP-aware sharing in BI– Eliminate redundancy with work sharing– Improve locality in query operators

• …

• But programmers are not multithreaded

Page 6: DB system design for new hardware and sciences

sciences

Page 7: DB system design for new hardware and sciences

Challenges: Complexity AND size

Alliez et al, INRIA, SIGGRAPH’05

Brain Mind Institute, EPFL

• Automate DB Design

• Computational Support

• Understanding Data

Page 8: DB system design for new hardware and sciences

Summary

• Challenge #1: exploit hardware– Parallelism, sharing maximized simultaneously– Infrastructure to parallel thinking&programming

• Challenge #2: serve sciences– Reduce complexity through abstraction– Manage large datasets on large computers

Page 9: DB system design for new hardware and sciences

“Multicore: This is the one which will have the biggest impact on us. We have never had a problem to solve like this. A breakthrough is needed in how applications are done on multicore devices.”

– Bill Gates

“It’s time we rethought some of the basics of computing. It’s scary and lots of fun at the same time.”

– Burton Smith