SlideShare a Scribd company logo
1 of 38
Download to read offline
Blosc

Sending data from memory to CPU (and back)	

faster than memcpy()
Francesc Alted

Software Architect

PyData London 2014	

February 22, 2014
About Me
• I am the creator of tools like PyTables,

Blosc, BLZ and maintainer of Numexpr.	


• I learnt the hard way that ‘premature
optimization is the root of all evil’.	


• Now I only humbly try to optimize if I

really need to and I just hope that Blosc is
not an example of ‘premature optimization’.
About Continuum Analytics

• Develop new ways on how data is stored,
computed, and visualized.	


• Provide open technologies for data
integration on a massive scale.	


• Provide software tools, training, and

integration/consulting services to
corporate, government, and educational
clients worldwide.
Overview
• Compressing faster than memcpy(). Really?	

• How that can be?

(The ‘Starving CPU’ problem)	


• How Blosc works.	

• Being faster than memcpy() means that my
programs would actually run faster?
Compressing Faster
than memcpy()
Interactive Session Starts
• If you 	

want to experiment with Blosc in
your own machine: 

http://www.blosc.org/materials/PyDataLondon-2014.tar.gz	


• blosc (blz too for later on) is required (both
are included in conda repository).
Open Questions
We have seen that, sometimes, Blosc can actually
be faster than memcpy(). Now:	

1. If compression takes way more CPU than
memcpy(), why Blosc can beat it?	

2. Does this mean that Blosc can actually
accelerate computations in real
scenarios?
“Across the industry, today’s chips are largely
able to execute code faster than we can feed
them with instructions and data.”	

!

– Richard Sites, after his article

“It’s The Memory, Stupid!”, 

Microprocessor Report, 10(10),1996

The Starving CPU
Problem
Memory Access Time
vs CPU Cycle Time
Book in
2009
The Status of CPU
Starvation in 2014
• Memory latency (~10 ns) is much slower

(between 100x and 250x) than processors.	


• Memory bandwidth (~15 GB/s) is

improving at a better rate than memory
latency, but it is also slower than
processors (between 30x and 100x).
Blosc Goals and
Implementation
Blosc: (de)compressing
faster than memcpy()

Transmission + decompression faster than direct transfer?
Taking Advantage of
Memory-CPU Gap
• Blosc is meant to discover redundancy in
data as fast as possible.	


• It comes with a series of fast compressors:
BloscLZ, LZ4, Snappy, LZ4HC and Zlib	


• Blosc is meant for speed, not for high
compression ratios.
Blosc Is All About
Efficiency
• Uses data blocks that fit in L1 or L2 caches
(better speed, less compression ratios).	


• Uses multithreading by default.	

• The shuffle filter uses SSE2 instructions in
modern Intel and AMD processors.
Blocking: Divide and
Conquer
Suffling: Improving the
Compression Ratio
The shuffling algorithm does not actually
compress the data; it rather changes the byte
order in the data stream:
Shuffling Caveat
• Shuffling usually produces better

compression ratios with numerical data,
except when it does not.	


• If you mind about the compression ratio, it
is worth to deactivate it and check (it is
active by default).	


• Will see an example on real data later on.
Blosc Performance:
Laptop back in 2005
Blosc Performance:
Desktop Computer in 2012
First Answer for Open
Questions
• Blosc data blocking optimizes the cache
behavior during memory access.	


• Additionally, it uses multithreading and
SIMD instructions.	


• Add these to the Starved CPU problem and

you have a good hint now on why Blosc can
beat memcpy().
How Compression
Works With Real Data?
The Need for
Compression
• Compression allows to store more data
using the same storage capacity.	


• Sure, it uses more CPU time to compress/
decompress data.	


• But, that actually means using more wall
clock time?
The Need for a
Compressed Container
• A compressed container is meant to store

data in compressed state and transparently
deliver it uncompressed.	


• That means that the user only perceives
that her dataset takes less memory.	


• Only less space? What about data access
speed?
Example of How Blosc Accelerates Genomics I/O:	

SeqDB (backed by Blosc)

Source: Howison, M. High-throughput compression of FASTQ data with SeqDB.
IEEE Transactions on Computational Biology and Bioinformatics.
Bloscpack (I)
• Command line interface and serialization
format for Blosc:	


!

$ blpk c data.dat

# compress

$ blpk d data.dat.blp

# decompress
Bloscpack (II)
• Very convenient for easily serializing your
in-memory NumPy datasets:	


>>> a = np.linspace(0, 1, 3e8)
>>> print a.size, a.dtype
300000000 float64
>>> bp.pack_ndarray_file(a, 'a.blp')
>>> b = bp.unpack_ndarray_file('a.blp')
>>> (a == b).all()
True
Yet Another Example: 	

BLZ	

• BLZ is a both a format and library that has

been designed as an efficient data container
for Big Data.	


• Blosc and Bloscpack are at the heart of it in
order to achieve high-speed compression/
decompression.	


• BLZ is one of the backends supported by
our nascent Blaze library.
Appending Data in
Large NumPy Objects
array to be enlarged

final array object
Copy!

new data to append

New memory	

allocation

• Normally a realloc() syscall will not succeed	

• Both memory areas have to exist simultaneously
Contiguous vs Chunked
NumPy container

BLZ container
chunk 1
chunk 2
.
.
.
chunk N

Contiguous memory

Discontiguous memory
Appending data in BLZ
array to be enlarged
chunk 1
chunk 2

new data to append

X
compress

final array object
chunk 1
chunk 2

new chunk

Only a small amount of data has to be compressed
The btable object in BLZ
Chunks

New row to append

• Columns are contiguous in memory	

• Chunks follow column order	

• Very efficient for querying (specially with a

large number of columns)
Second Interactive
Session: BLZ and Blosc
on a Real Dataset
Second Hint for Open
Questions	

Blosc usage in BLZ means not only less storage
usage (~15x-40x reduction for the real life data
shown), but almost the same access time to
the data (~2x-10x slowdown).	

(Still need to address implementation details for
getting better performance)
Summary
• Blosc, being able to transfer data faster than
memcpy(), has enormous implications on
data management.	


• It is well suited not only for saving memory,
but for allowing close performance to
typical uncompressed data containers.	


• It works well not only for synthetic data,
but also for real-life datasets.
References
• Blosc: http://www.blosc.org	

• Bloscpack: https://github.com/Blosc/bloscpack	

• BLZ: http://blz.pydata.org
“Across the industry, today’s chips are largely able to execute code
faster than we can feed them with instructions and data. There are no
longer performance bottlenecks in the floating-point multiplier or in
having only a single integer unit. The real design action is in memory
subsystems— caches, buses, bandwidth, and latency.”	

!

“Over the coming decade, memory subsystem design will be the only
important design issue for microprocessors.”	

!

– Richard Sites, after his article “It’s The Memory, Stupid!”,
Microprocessor Report, 10(10),1996
“Over this decade (2010-2020), memory subsystem optimization
will be (almost) the only important design issue for improving
performance.”	

– Me :)
Thank you!

More Related Content

What's hot

全社情報共有サイトへの Alfresco Community 5 導入事例紹介 - 第27回Alfresco勉強会
全社情報共有サイトへのAlfresco Community 5 導入事例紹介 - 第27回Alfresco勉強会全社情報共有サイトへのAlfresco Community 5 導入事例紹介 - 第27回Alfresco勉強会
全社情報共有サイトへの Alfresco Community 5 導入事例紹介 - 第27回Alfresco勉強会
Ryota Watabe
 

What's hot (20)

全社情報共有サイトへの Alfresco Community 5 導入事例紹介 - 第27回Alfresco勉強会
全社情報共有サイトへのAlfresco Community 5 導入事例紹介 - 第27回Alfresco勉強会全社情報共有サイトへのAlfresco Community 5 導入事例紹介 - 第27回Alfresco勉強会
全社情報共有サイトへの Alfresco Community 5 導入事例紹介 - 第27回Alfresco勉強会
 
How We Made Scylla Maintenance Easier, Safer and Faster
How We Made Scylla Maintenance Easier, Safer and FasterHow We Made Scylla Maintenance Easier, Safer and Faster
How We Made Scylla Maintenance Easier, Safer and Faster
 
PostGIS 初入門應用
PostGIS 初入門應用PostGIS 初入門應用
PostGIS 初入門應用
 
PHP でバイナリ変換プログラミング
PHP でバイナリ変換プログラミングPHP でバイナリ変換プログラミング
PHP でバイナリ変換プログラミング
 
続・広く知ってほしいDNSのこと
続・広く知ってほしいDNSのこと続・広く知ってほしいDNSのこと
続・広く知ってほしいDNSのこと
 
2019 年後半 海外動画技術動向
2019 年後半 海外動画技術動向2019 年後半 海外動画技術動向
2019 年後半 海外動画技術動向
 
OpenZFS novel algorithms: snapshots, space allocation, RAID-Z - Matt Ahrens
OpenZFS novel algorithms: snapshots, space allocation, RAID-Z - Matt AhrensOpenZFS novel algorithms: snapshots, space allocation, RAID-Z - Matt Ahrens
OpenZFS novel algorithms: snapshots, space allocation, RAID-Z - Matt Ahrens
 
EBPF and Linux Networking
EBPF and Linux NetworkingEBPF and Linux Networking
EBPF and Linux Networking
 
仮想環境の設計手法
仮想環境の設計手法仮想環境の設計手法
仮想環境の設計手法
 
MySQLと組み合わせて始める全文検索プロダクト"elasticsearch"
MySQLと組み合わせて始める全文検索プロダクト"elasticsearch"MySQLと組み合わせて始める全文検索プロダクト"elasticsearch"
MySQLと組み合わせて始める全文検索プロダクト"elasticsearch"
 
eBPF Workshop
eBPF WorkshopeBPF Workshop
eBPF Workshop
 
IBM DS8880 and IBM Z - Integrated by Design
IBM DS8880 and IBM Z - Integrated by DesignIBM DS8880 and IBM Z - Integrated by Design
IBM DS8880 and IBM Z - Integrated by Design
 
さくらのVPS で IPv4 over IPv6ルータの構築
さくらのVPS で IPv4 over IPv6ルータの構築さくらのVPS で IPv4 over IPv6ルータの構築
さくらのVPS で IPv4 over IPv6ルータの構築
 
How to Choose the Right Database for Your Workloads
How to Choose the Right Database for Your WorkloadsHow to Choose the Right Database for Your Workloads
How to Choose the Right Database for Your Workloads
 
MQTTとAMQPと.NET
MQTTとAMQPと.NETMQTTとAMQPと.NET
MQTTとAMQPと.NET
 
Using The Mysql Binary Log As A Change Stream
Using The Mysql Binary Log As A Change StreamUsing The Mysql Binary Log As A Change Stream
Using The Mysql Binary Log As A Change Stream
 
よろしい、ならばMicro-ORMだ
よろしい、ならばMicro-ORMだよろしい、ならばMicro-ORMだ
よろしい、ならばMicro-ORMだ
 
Oracle RAC 19c and Later - Best Practices #OOWLON
Oracle RAC 19c and Later - Best Practices #OOWLONOracle RAC 19c and Later - Best Practices #OOWLON
Oracle RAC 19c and Later - Best Practices #OOWLON
 
Building Thick Clients with Tower in Rust
Building Thick Clients with Tower in RustBuilding Thick Clients with Tower in Rust
Building Thick Clients with Tower in Rust
 
MySQLとPostgreSQLと日本語全文検索 - Azure DatabaseでMroonga・PGroongaを使いたいですよね!?
MySQLとPostgreSQLと日本語全文検索 - Azure DatabaseでMroonga・PGroongaを使いたいですよね!?MySQLとPostgreSQLと日本語全文検索 - Azure DatabaseでMroonga・PGroongaを使いたいですよね!?
MySQLとPostgreSQLと日本語全文検索 - Azure DatabaseでMroonga・PGroongaを使いたいですよね!?
 

Similar to Blosc Talk by Francesc Alted from PyData London 2014

Ceph Day Amsterdam 2015: Measuring and predicting performance of Ceph clusters
Ceph Day Amsterdam 2015: Measuring and predicting performance of Ceph clusters Ceph Day Amsterdam 2015: Measuring and predicting performance of Ceph clusters
Ceph Day Amsterdam 2015: Measuring and predicting performance of Ceph clusters
Ceph Community
 
Deployment Strategy
Deployment StrategyDeployment Strategy
Deployment Strategy
MongoDB
 

Similar to Blosc Talk by Francesc Alted from PyData London 2014 (20)

PyData Paris 2015 - Closing keynote Francesc Alted
PyData Paris 2015 - Closing keynote Francesc AltedPyData Paris 2015 - Closing keynote Francesc Alted
PyData Paris 2015 - Closing keynote Francesc Alted
 
It's the memory, stupid! CodeJam 2014
It's the memory, stupid!  CodeJam 2014It's the memory, stupid!  CodeJam 2014
It's the memory, stupid! CodeJam 2014
 
Elements of cache design
Elements of cache designElements of cache design
Elements of cache design
 
Ceph Day Amsterdam 2015: Measuring and predicting performance of Ceph clusters
Ceph Day Amsterdam 2015: Measuring and predicting performance of Ceph clusters Ceph Day Amsterdam 2015: Measuring and predicting performance of Ceph clusters
Ceph Day Amsterdam 2015: Measuring and predicting performance of Ceph clusters
 
Limitations of memory system performance
Limitations of memory system performanceLimitations of memory system performance
Limitations of memory system performance
 
computer-memory
computer-memorycomputer-memory
computer-memory
 
Data warehouse 26 exploiting parallel technologies
Data warehouse  26 exploiting parallel technologiesData warehouse  26 exploiting parallel technologies
Data warehouse 26 exploiting parallel technologies
 
Cache Memory.pptx
Cache Memory.pptxCache Memory.pptx
Cache Memory.pptx
 
Training Webinar: Enterprise application performance with distributed caching
Training Webinar: Enterprise application performance with distributed cachingTraining Webinar: Enterprise application performance with distributed caching
Training Webinar: Enterprise application performance with distributed caching
 
Real time database compression optimization using iterative length compressio...
Real time database compression optimization using iterative length compressio...Real time database compression optimization using iterative length compressio...
Real time database compression optimization using iterative length compressio...
 
REAL TIME DATABASE COMPRESSION OPTIMIZATION USING ITERATIVE LENGTH COMPRESSIO...
REAL TIME DATABASE COMPRESSION OPTIMIZATION USING ITERATIVE LENGTH COMPRESSIO...REAL TIME DATABASE COMPRESSION OPTIMIZATION USING ITERATIVE LENGTH COMPRESSIO...
REAL TIME DATABASE COMPRESSION OPTIMIZATION USING ITERATIVE LENGTH COMPRESSIO...
 
COA notes
COA notesCOA notes
COA notes
 
High Performance With Java
High Performance With JavaHigh Performance With Java
High Performance With Java
 
Scaling up Machine Learning Algorithms for Classification
Scaling up Machine Learning Algorithms for ClassificationScaling up Machine Learning Algorithms for Classification
Scaling up Machine Learning Algorithms for Classification
 
I-Sieve: An inline High Performance Deduplication System Used in cloud storage
I-Sieve: An inline High Performance Deduplication System Used in cloud storageI-Sieve: An inline High Performance Deduplication System Used in cloud storage
I-Sieve: An inline High Performance Deduplication System Used in cloud storage
 
Deployment Strategies
Deployment StrategiesDeployment Strategies
Deployment Strategies
 
Cache coherence ppt
Cache coherence pptCache coherence ppt
Cache coherence ppt
 
5 Pitfalls to Avoid with MongoDB
5 Pitfalls to Avoid with MongoDB5 Pitfalls to Avoid with MongoDB
5 Pitfalls to Avoid with MongoDB
 
Deployment Strategy
Deployment StrategyDeployment Strategy
Deployment Strategy
 
The Fundamental Characteristics of Storage concepts for DBAs
The Fundamental Characteristics of Storage concepts for DBAsThe Fundamental Characteristics of Storage concepts for DBAs
The Fundamental Characteristics of Storage concepts for DBAs
 

More from PyData

More from PyData (20)

Michal Mucha: Build and Deploy an End-to-end Streaming NLP Insight System | P...
Michal Mucha: Build and Deploy an End-to-end Streaming NLP Insight System | P...Michal Mucha: Build and Deploy an End-to-end Streaming NLP Insight System | P...
Michal Mucha: Build and Deploy an End-to-end Streaming NLP Insight System | P...
 
Unit testing data with marbles - Jane Stewart Adams, Leif Walsh
Unit testing data with marbles - Jane Stewart Adams, Leif WalshUnit testing data with marbles - Jane Stewart Adams, Leif Walsh
Unit testing data with marbles - Jane Stewart Adams, Leif Walsh
 
The TileDB Array Data Storage Manager - Stavros Papadopoulos, Jake Bolewski
The TileDB Array Data Storage Manager - Stavros Papadopoulos, Jake BolewskiThe TileDB Array Data Storage Manager - Stavros Papadopoulos, Jake Bolewski
The TileDB Array Data Storage Manager - Stavros Papadopoulos, Jake Bolewski
 
Using Embeddings to Understand the Variance and Evolution of Data Science... ...
Using Embeddings to Understand the Variance and Evolution of Data Science... ...Using Embeddings to Understand the Variance and Evolution of Data Science... ...
Using Embeddings to Understand the Variance and Evolution of Data Science... ...
 
Deploying Data Science for Distribution of The New York Times - Anne Bauer
Deploying Data Science for Distribution of The New York Times - Anne BauerDeploying Data Science for Distribution of The New York Times - Anne Bauer
Deploying Data Science for Distribution of The New York Times - Anne Bauer
 
Graph Analytics - From the Whiteboard to Your Toolbox - Sam Lerma
Graph Analytics - From the Whiteboard to Your Toolbox - Sam LermaGraph Analytics - From the Whiteboard to Your Toolbox - Sam Lerma
Graph Analytics - From the Whiteboard to Your Toolbox - Sam Lerma
 
Do Your Homework! Writing tests for Data Science and Stochastic Code - David ...
Do Your Homework! Writing tests for Data Science and Stochastic Code - David ...Do Your Homework! Writing tests for Data Science and Stochastic Code - David ...
Do Your Homework! Writing tests for Data Science and Stochastic Code - David ...
 
RESTful Machine Learning with Flask and TensorFlow Serving - Carlo Mazzaferro
RESTful Machine Learning with Flask and TensorFlow Serving - Carlo MazzaferroRESTful Machine Learning with Flask and TensorFlow Serving - Carlo Mazzaferro
RESTful Machine Learning with Flask and TensorFlow Serving - Carlo Mazzaferro
 
Mining dockless bikeshare and dockless scootershare trip data - Stefanie Brod...
Mining dockless bikeshare and dockless scootershare trip data - Stefanie Brod...Mining dockless bikeshare and dockless scootershare trip data - Stefanie Brod...
Mining dockless bikeshare and dockless scootershare trip data - Stefanie Brod...
 
Avoiding Bad Database Surprises: Simulation and Scalability - Steven Lott
Avoiding Bad Database Surprises: Simulation and Scalability - Steven LottAvoiding Bad Database Surprises: Simulation and Scalability - Steven Lott
Avoiding Bad Database Surprises: Simulation and Scalability - Steven Lott
 
Words in Space - Rebecca Bilbro
Words in Space - Rebecca BilbroWords in Space - Rebecca Bilbro
Words in Space - Rebecca Bilbro
 
End-to-End Machine learning pipelines for Python driven organizations - Nick ...
End-to-End Machine learning pipelines for Python driven organizations - Nick ...End-to-End Machine learning pipelines for Python driven organizations - Nick ...
End-to-End Machine learning pipelines for Python driven organizations - Nick ...
 
Pydata beautiful soup - Monica Puerto
Pydata beautiful soup - Monica PuertoPydata beautiful soup - Monica Puerto
Pydata beautiful soup - Monica Puerto
 
1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...
1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...
1D Convolutional Neural Networks for Time Series Modeling - Nathan Janos, Jef...
 
Extending Pandas with Custom Types - Will Ayd
Extending Pandas with Custom Types - Will AydExtending Pandas with Custom Types - Will Ayd
Extending Pandas with Custom Types - Will Ayd
 
Measuring Model Fairness - Stephen Hoover
Measuring Model Fairness - Stephen HooverMeasuring Model Fairness - Stephen Hoover
Measuring Model Fairness - Stephen Hoover
 
What's the Science in Data Science? - Skipper Seabold
What's the Science in Data Science? - Skipper SeaboldWhat's the Science in Data Science? - Skipper Seabold
What's the Science in Data Science? - Skipper Seabold
 
Applying Statistical Modeling and Machine Learning to Perform Time-Series For...
Applying Statistical Modeling and Machine Learning to Perform Time-Series For...Applying Statistical Modeling and Machine Learning to Perform Time-Series For...
Applying Statistical Modeling and Machine Learning to Perform Time-Series For...
 
Solving very simple substitution ciphers algorithmically - Stephen Enright-Ward
Solving very simple substitution ciphers algorithmically - Stephen Enright-WardSolving very simple substitution ciphers algorithmically - Stephen Enright-Ward
Solving very simple substitution ciphers algorithmically - Stephen Enright-Ward
 
The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...
The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...
The Face of Nanomaterials: Insightful Classification Using Deep Learning - An...
 

Recently uploaded

Finding Java's Hidden Performance Traps @ DevoxxUK 2024
Finding Java's Hidden Performance Traps @ DevoxxUK 2024Finding Java's Hidden Performance Traps @ DevoxxUK 2024
Finding Java's Hidden Performance Traps @ DevoxxUK 2024
Victor Rentea
 

Recently uploaded (20)

Mcleodganj Call Girls 🥰 8617370543 Service Offer VIP Hot Model
Mcleodganj Call Girls 🥰 8617370543 Service Offer VIP Hot ModelMcleodganj Call Girls 🥰 8617370543 Service Offer VIP Hot Model
Mcleodganj Call Girls 🥰 8617370543 Service Offer VIP Hot Model
 
[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf
 
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
 
Understanding the FAA Part 107 License ..
Understanding the FAA Part 107 License ..Understanding the FAA Part 107 License ..
Understanding the FAA Part 107 License ..
 
Vector Search -An Introduction in Oracle Database 23ai.pptx
Vector Search -An Introduction in Oracle Database 23ai.pptxVector Search -An Introduction in Oracle Database 23ai.pptx
Vector Search -An Introduction in Oracle Database 23ai.pptx
 
Finding Java's Hidden Performance Traps @ DevoxxUK 2024
Finding Java's Hidden Performance Traps @ DevoxxUK 2024Finding Java's Hidden Performance Traps @ DevoxxUK 2024
Finding Java's Hidden Performance Traps @ DevoxxUK 2024
 
CNIC Information System with Pakdata Cf In Pakistan
CNIC Information System with Pakdata Cf In PakistanCNIC Information System with Pakdata Cf In Pakistan
CNIC Information System with Pakdata Cf In Pakistan
 
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, AdobeApidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
 
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdfRising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
 
"I see eyes in my soup": How Delivery Hero implemented the safety system for ...
"I see eyes in my soup": How Delivery Hero implemented the safety system for ..."I see eyes in my soup": How Delivery Hero implemented the safety system for ...
"I see eyes in my soup": How Delivery Hero implemented the safety system for ...
 
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot TakeoffStrategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
 
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWEREMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
 
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
 
Apidays New York 2024 - APIs in 2030: The Risk of Technological Sleepwalk by ...
Apidays New York 2024 - APIs in 2030: The Risk of Technological Sleepwalk by ...Apidays New York 2024 - APIs in 2030: The Risk of Technological Sleepwalk by ...
Apidays New York 2024 - APIs in 2030: The Risk of Technological Sleepwalk by ...
 
MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024
 
Six Myths about Ontologies: The Basics of Formal Ontology
Six Myths about Ontologies: The Basics of Formal OntologySix Myths about Ontologies: The Basics of Formal Ontology
Six Myths about Ontologies: The Basics of Formal Ontology
 
DBX First Quarter 2024 Investor Presentation
DBX First Quarter 2024 Investor PresentationDBX First Quarter 2024 Investor Presentation
DBX First Quarter 2024 Investor Presentation
 
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin WoodPolkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
 
Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...
Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...
Web Form Automation for Bonterra Impact Management (fka Social Solutions Apri...
 
Artificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : UncertaintyArtificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : Uncertainty
 

Blosc Talk by Francesc Alted from PyData London 2014

  • 1. Blosc Sending data from memory to CPU (and back) faster than memcpy() Francesc Alted
 Software Architect
 PyData London 2014 February 22, 2014
  • 2. About Me • I am the creator of tools like PyTables, Blosc, BLZ and maintainer of Numexpr. • I learnt the hard way that ‘premature optimization is the root of all evil’. • Now I only humbly try to optimize if I really need to and I just hope that Blosc is not an example of ‘premature optimization’.
  • 3. About Continuum Analytics • Develop new ways on how data is stored, computed, and visualized. • Provide open technologies for data integration on a massive scale. • Provide software tools, training, and integration/consulting services to corporate, government, and educational clients worldwide.
  • 4. Overview • Compressing faster than memcpy(). Really? • How that can be?
 (The ‘Starving CPU’ problem) • How Blosc works. • Being faster than memcpy() means that my programs would actually run faster?
  • 6. Interactive Session Starts • If you want to experiment with Blosc in your own machine: 
 http://www.blosc.org/materials/PyDataLondon-2014.tar.gz • blosc (blz too for later on) is required (both are included in conda repository).
  • 7. Open Questions We have seen that, sometimes, Blosc can actually be faster than memcpy(). Now: 1. If compression takes way more CPU than memcpy(), why Blosc can beat it? 2. Does this mean that Blosc can actually accelerate computations in real scenarios?
  • 8. “Across the industry, today’s chips are largely able to execute code faster than we can feed them with instructions and data.” ! – Richard Sites, after his article
 “It’s The Memory, Stupid!”, 
 Microprocessor Report, 10(10),1996 The Starving CPU Problem
  • 9. Memory Access Time vs CPU Cycle Time
  • 11. The Status of CPU Starvation in 2014 • Memory latency (~10 ns) is much slower (between 100x and 250x) than processors. • Memory bandwidth (~15 GB/s) is improving at a better rate than memory latency, but it is also slower than processors (between 30x and 100x).
  • 13. Blosc: (de)compressing faster than memcpy() Transmission + decompression faster than direct transfer?
  • 14. Taking Advantage of Memory-CPU Gap • Blosc is meant to discover redundancy in data as fast as possible. • It comes with a series of fast compressors: BloscLZ, LZ4, Snappy, LZ4HC and Zlib • Blosc is meant for speed, not for high compression ratios.
  • 15. Blosc Is All About Efficiency • Uses data blocks that fit in L1 or L2 caches (better speed, less compression ratios). • Uses multithreading by default. • The shuffle filter uses SSE2 instructions in modern Intel and AMD processors.
  • 17. Suffling: Improving the Compression Ratio The shuffling algorithm does not actually compress the data; it rather changes the byte order in the data stream:
  • 18. Shuffling Caveat • Shuffling usually produces better compression ratios with numerical data, except when it does not. • If you mind about the compression ratio, it is worth to deactivate it and check (it is active by default). • Will see an example on real data later on.
  • 21. First Answer for Open Questions • Blosc data blocking optimizes the cache behavior during memory access. • Additionally, it uses multithreading and SIMD instructions. • Add these to the Starved CPU problem and you have a good hint now on why Blosc can beat memcpy().
  • 23. The Need for Compression • Compression allows to store more data using the same storage capacity. • Sure, it uses more CPU time to compress/ decompress data. • But, that actually means using more wall clock time?
  • 24. The Need for a Compressed Container • A compressed container is meant to store data in compressed state and transparently deliver it uncompressed. • That means that the user only perceives that her dataset takes less memory. • Only less space? What about data access speed?
  • 25. Example of How Blosc Accelerates Genomics I/O: SeqDB (backed by Blosc) Source: Howison, M. High-throughput compression of FASTQ data with SeqDB. IEEE Transactions on Computational Biology and Bioinformatics.
  • 26. Bloscpack (I) • Command line interface and serialization format for Blosc: ! $ blpk c data.dat # compress $ blpk d data.dat.blp # decompress
  • 27. Bloscpack (II) • Very convenient for easily serializing your in-memory NumPy datasets: >>> a = np.linspace(0, 1, 3e8) >>> print a.size, a.dtype 300000000 float64 >>> bp.pack_ndarray_file(a, 'a.blp') >>> b = bp.unpack_ndarray_file('a.blp') >>> (a == b).all() True
  • 28. Yet Another Example: BLZ • BLZ is a both a format and library that has been designed as an efficient data container for Big Data. • Blosc and Bloscpack are at the heart of it in order to achieve high-speed compression/ decompression. • BLZ is one of the backends supported by our nascent Blaze library.
  • 29. Appending Data in Large NumPy Objects array to be enlarged final array object Copy! new data to append New memory allocation • Normally a realloc() syscall will not succeed • Both memory areas have to exist simultaneously
  • 30. Contiguous vs Chunked NumPy container BLZ container chunk 1 chunk 2 . . . chunk N Contiguous memory Discontiguous memory
  • 31. Appending data in BLZ array to be enlarged chunk 1 chunk 2 new data to append X compress final array object chunk 1 chunk 2 new chunk Only a small amount of data has to be compressed
  • 32. The btable object in BLZ Chunks New row to append • Columns are contiguous in memory • Chunks follow column order • Very efficient for querying (specially with a
 large number of columns)
  • 33. Second Interactive Session: BLZ and Blosc on a Real Dataset
  • 34. Second Hint for Open Questions Blosc usage in BLZ means not only less storage usage (~15x-40x reduction for the real life data shown), but almost the same access time to the data (~2x-10x slowdown). (Still need to address implementation details for getting better performance)
  • 35. Summary • Blosc, being able to transfer data faster than memcpy(), has enormous implications on data management. • It is well suited not only for saving memory, but for allowing close performance to typical uncompressed data containers. • It works well not only for synthetic data, but also for real-life datasets.
  • 36. References • Blosc: http://www.blosc.org • Bloscpack: https://github.com/Blosc/bloscpack • BLZ: http://blz.pydata.org
  • 37. “Across the industry, today’s chips are largely able to execute code faster than we can feed them with instructions and data. There are no longer performance bottlenecks in the floating-point multiplier or in having only a single integer unit. The real design action is in memory subsystems— caches, buses, bandwidth, and latency.” ! “Over the coming decade, memory subsystem design will be the only important design issue for microprocessors.” ! – Richard Sites, after his article “It’s The Memory, Stupid!”, Microprocessor Report, 10(10),1996 “Over this decade (2010-2020), memory subsystem optimization will be (almost) the only important design issue for improving performance.” – Me :)