Andy Pavlo // Carnegie Mellon University // Spring 2016
Lecture #12 – Join Algorithms (Sorting)
DATABASE
SYSTEMS
15-721
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Guidelines and tips
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Community
Ask the community
Ask the community for help and clear up your study doubts
University Rankings
Discover the best universities in your country according to Docsity users
Free resources
Our save-the-student-ebooks!
Download our free guides on studying techniques, anxiety management strategies, and thesis advice from Docsity tutors
Sorting and Join Algorithms in Database Systems: Lecture 12 by Andy Pavlo at CMU, Study notes of Algorithms and Programming
A lecture note from Carnegie Mellon University (CMU) course 15-721 (Spring 2016) on Database Systems. The lecture, titled 'Join Algorithms (Sorting)', covers the topic of sorting and merge join algorithms in the context of database systems. agendas, diagrams, and references to related research papers.
Typology: Study notes
2021/2022
1 / 123
This page cannot be seen from the preview
Don't miss anything!
Related documents
Partial preview of the text
Download Sorting and Join Algorithms in Database Systems: Lecture 12 by Andy Pavlo at CMU and more Study notes Algorithms and Programming in PDF only on Docsity!
Andy Pavlo // Carnegie Mellon University // Spring 2016
Lecture #12 – Join Algorithms (Sorting)
DATABASE SYSTEMS
15-
TO DAY ’ S AG E N DA
Background SIMD
Parallel Sort-Merge Join Evaluation
Hate Mail
S O R T- M E R G E J O I N ( R ⨝ S )
Relation R Relation S
S O R T- M E R G E J O I N ( R ⨝ S )
Relation R Relation S
SORT!
SORT!
S O R T- M E R G E J O I N ( R ⨝ S )
Relation R Relation S
SORT!^ ⨝
SORT!
MERGE!
S O R T- M E R G E J O I N ( R ⨝ S )
Relation R Relation S
SORT!^ ⨝
SORT!
MERGE!
I N - M E M O R Y J O I N S
→ → Hashing is faster than Sort-Merge.Sort-Merge will be faster with wider SIMD.
SORT VS. HASH REVISITED: FAST JOIN IMPLEMENTATION ON MODERN MULTI VLDB 2009 - CORE CPUS
→ Sort-Merge is already faster,even without SIMD.
MASSIVELY JOINS IN MAIN MEMORY MULTI CORE DATABASE SYSTEMS VLDB 2012 PARALLEL SORT-MERGE-
→ New optimizations andresults for Radix Hash Join.
MAIN MULTI UNDERLYING HARDWARE ICDE 2013 - MEMORY HASH JOINS ON-CORE CPUS: TUNING TO THE
Source: Cagri Balkesen
S I N G L E I N S T R U C T I O N , M U LT I P L E D ATA
A class of CPU instructions that allow theprocessor to perform the same operation on
multiple data points simultaneously.
Both current AMD and Intel CPUs have ISA andmicroarchitecture support SIMD operations.
→ MMX, 3DNow!, SSE, SSE2, SSE3, SSE4, AVX
S I M D E X A M P L E
X + Y = Z
n n xn y n
xx yy y
yy x
x ... x (^21) ... (^2121) ...^12
Z
S I M D E X A M P L E
X + Y = Z
n n xn y n
xx yy y
yy x
x ... x (^21) ... (^2121) ...^12
X for Z[i] = X[i] + Y[i]; (i=0; i<n; i++) { }
Y
Z
S I M D E X A M P L E
X + Y = Z
n n xn y n
xx yy y
yy x
x ... x (^21) ... (^2121) ...^12
X
SISD
for Z[i] = X[i] + Y[i]; (i=0; i<n; i++) {^ + }
Y
Z
S I M D E X A M P L E
X + Y = Z
n n xn y n
xx yy y
yy x
x ... x (^21) ... (^2121) ...^12
X
SISD
for Z[i] = X[i] + Y[i]; (i=0; i<n; i++) {^ + }
Y
Z
S I M D E X A M P L E
X + Y = Z
n n xn y n
xx yy y
yy x
x ... x (^21) ... (^2121) ...^12
X for Z[i] = X[i] + Y[i]; (i=0; i<n; i++) { }
Y
SIMD
Z
S I M D E X A M P L E
X + Y = Z
n n xn y n
xx yy y
yy x
x ... x (^21) ... (^2121) ...^12
X for Z[i] = X[i] + Y[i]; (i=0; i<n; i++) { }
Y