Intel Developer Forum: Taming the Big Data Tsunami
using Intel® Architecture by Clive D’Souza, Solutions Architect, Intel Corporation and
Dhruv Bansal, Chief Science Officer, Infochimps
Gen AI in Business - Global Trends Report 2024.pdf
Taming the Big Data Tsunami using Intel Architecture
1. Taming the Big Data Tsunami
using Intel® Architecture
Clive D’Souza, Solutions Architect, Intel Corporation
Dhruv Bansal, Chief Science Officer, Infochimps
DATS004
2. Agenda
• What is Big Data?
• Why does Big Data matter?
• How can Intel®
Architecture help
• Summary
2
3. Big Data Tsunami
Between the birth of the world and 2003,
there were five Exabyte of information
created. We now create five Exabyte
every two days
180,000 Eric Schmidt
160,000
Content Depots – Massive/
Over 24 Petabytes
Exponential Growth
140,000 Unstructured
Data processed by
Enterprise Hosting Services Google* every day in 2011
120,000
100,000
Four billion
Traditional Unstructured Pieces of content shared
80,000
on Facebook* every day by
July 2011
60,000
Traditional Structured
Data 250 Million
40,000 Tweets per day in October
Growth
2011
Linear
20,000
5.5 million
0 Legitimate emails sent
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
every second in 2011
2.7 Zetabytes of data in 2012, 15 billion connected
devices by 2015 !!!
Source: IDC, 2011 Worldwide Enterprise Storage Systems 2011–2015 Forecast Update.
3 Worldwide Enterprise Storage Consumption Capacity Shipped by Model, 2006–2015 (PB)
4. Big Data — Traits
Big Data Decrees
Volume • Speed is everything!
• Use diverse data
• Data never gets stale
Velocity Value
Variety • Data growth will be
exponential
• Big Data is real
Core Tenants • Transformational to business
Unstructured datasets whose Volume, Variety and
Velocity is beyond the ability of typical database
software tools to capture, store, manage and analyze†
4 †Big data: The next frontier for innovation, competition, and productivity”, McKinsey Global Institute
5. Big Data — Flow
Data Analytics
Data - Query
Enabled
Data Curation
Data
Aggregation
Data
Ingestion
Compute/Network/IO/Storage-Intensive
5
6. Agenda
• What is Big Data?
• Why does Big Data matter?
• How can Intel®
Architecture help
• Summary
6
7. Our Problem – Which 5K?
• Don’t know the future value of
today’s data
• We cannot connect the dots we do
not yet have
• The old collect, winnow, dissemble
model fails spectacularly in the Big
Data world
The “5K” is different for everybody!
Image used with permission from Author
7
8. Intelligent 5K = Big Money!
US health care Global personal
• $300 billion value per location data
year • $100 billion+ revenue
• ~0.7 percent annual for service providers
productivity growth • Up to $700 billion value
Manufacturing to end users
• Up to 50 percent
decrease in product
development, assembly
costs
• Up to seven percent
reduction in working
US retail capital
• 60+ percent increase in Europe public sector
net margin possible administration
• 0.5-1.0 percent annual • €250 billion value per
productivity growth year
• ~0.5 percent annual
productivity growth
8 Source: Mckinsey, 2011
10. Growing Pains…
• IT growth outpaced by Big
Data growth
• Unparalleled data complexity
• Need for speed – race to the
bottom!
• Workload management
• Data access, data silos, data
quality, data security
• Shortage of data scientists
• New domain – not easy to
implement
Big Data solutions will transform IT
10
11. Big Data Means More Than
Hadoop*
[Many people’s] understanding of “Hadoop” is like my
understanding of “tango”: I know the word, I know one
when I see one, but I can’t dance for ***.
Jeffrey Eisenberg
Hadoop* Ecosystem
• Java*
• Multi-language
• I/O-bound • Databases
• Batch, map/reduce,
historical
• Web
• Realtime
Big Data hardware needs more than I/O
11 About Infochimps* www.infochimps.com
12. Full Data Stack
Overview
Internal DBs & Data
Appliances
Data Storage
Public Integration (Analytic Hadoop*
Data Sensors
(ETL and DBs and
(Batch/Historical
Analytics)
streaming) filesystem)
Rich Media
CRM
ERM
POS Stream
Web Processing
(Real-Time
Logs Mobile Analytics)
System
Logs Documents
12
13. Full Data Stack
Database
Sources
Sqoop*
Bulk Load Structured Hadoop*
Datastores like SQL Hbase*
or Primary Analytics Bulk/Large-Scale
HDFS* Datastore Processing Engine
Streaming
Sources Flume*
Collect and Process Streaming
or Fast-Changing Data
Elasticsearch* Mongo* or
Search Engine and MySQL* Tableau*
API Reporting LogiXML*
Aggregates Custom App
etc.
DataViz*
Local SAS*, R*, Web Servers
Stata*, etc.
Disk Statistical
13
Packages
14. Demonstration
When are two time series correlated?
AAPL
14
15. Demonstration
Q: Traffic to which Wikipedia* articles
is correlated with the price of AAPL?
AAPL
Source Data:
• Web traffic logs (Wikipedia, 3 mos.)
• S&P 500* Stock Prices
15
16. Demonstration
Tentative Answer: Traffic to
articles about music, television,
and video games are directly
correlated with AAPL’s stock price.
Bonus: Also Jack
AAPL
Dorsey, CEO of
Square!
16
18. Agenda
• What is Big Data?
• Why does Big Data matter?
• How can Intel®
Architecture help
• Summary
18
19. Intelligent Data Center
HPC & IOPS/TB
Decision Optimized
Support
Dedicated Premium
Edg
e/M Servers Storage
2M
VPN or LAN Low- SSD
Latency, “Centralized”
Compute Proximity Unified Storage
WWW Storage Network
High-
Virtualized
Servers NVM Capacity
Storage
HDD
IT/Web/Application
Development
$/TB
Infrastructures
Optimized
19
20. Intel® Xeon® Processor = Heart
of the Intelligent Data Center
• Integrated PCI Express* Gen 3.0
• Intel® Hyper-Threading
Technology, two Threads/Core
• Shared Last Level Cache, 2.5 MB/
Core
• Higher memory bandwidth with
DDR3
• Integrated Memory Controller
• PCIe Non-Transparent Bridge
• Asynchronous DRAM self-refresh
(ADR)
• Intel® QuickData Technology
Direct Memory Access
Intel® Xeon® powers Big Data compute
20
21. Intelligent Storage Optimizations
De-duplication Real Time Compression
BEFORE DE-DUPLICATIONAFTER
95% smaller backup1 Up to 80% data reduction2
Intelligent Tiering Thin Provisioning
TRADITIONAL ALLOCATION THIN PROVISIONING
ALLOCATED BUT FREE
APPLI 3 USED
SYSTEM-WIDE
ALLOCATED BUT FREE CAPACITY
APPLI 2
RESERVED
USED
APPLI 3
ALLOCATED BUT FREE
APPLI 2
APPLI 1
USED APPLI 1
Up to 80% reduction in Up to 25% reduction in
disk expenses3 annual storage CapEx
growth4
1 IBM storage simulcast, November 9, 2011
2 BM storage simulcast, November 9, 2011
3 Dell “Fluid Data Storage: Driving Flexibility in the Data Center”, February 2011
21 4 Intel IT study “Solving Intel IT’s Data Storage Growth Challenges
22. New Memory Hierarchies —
Non-Volatile Memory
Time spent by application
in CPU vs. IO
Intel® Solid-State Drive
Application 910 Series
CPU Processing
• Enhanced Performance
- Sequential R/W: 2.0/1.0 GB/s
- Random R/W: 180/75 KIOPS
Timeline
SW 10µs
- Latency R/W: 65/65µs
IO
Processing
NVM 65µs • High Endurance 25nm HET MLC
- 10x drive writes/day for five years
- 30x endurance over standard MLC
due to improved write amplitude
and NAND management
CPU
Processing
Reduction of software latency dramatically increases
application IOPS as NVM latency decreases
22
23. Intel® Integrated IO Technology
Inbound Flow (Rx) Outbound Flow (Tx)
Core Reads Data
2 Intel Xeon® 1 Intel Xeon
Core creates buffer
Intel Processor E5-2600 Intel Processor E5-2600
for I/O device to read,
QPI 1 QPI 1 putting data in cache
Intel (cache line allocated)
QPI 2 CORE 1 CORE 2 Intel CORE 1 CORE 2
QPI 2
CORE 3 CORE 4 CORE 3 CORE 4
CORE 5 CORE 6 CORE 5 CORE 6
CORE 7 CORE 83 CORE 7 CORE 8
LLC Data to Core 3
Tx Packet
Data to I/O
CACHE IOC CACHE
1
DMA Write directly
To “IO allocated” LLC
No
No Memory
Memory Transactions
Transactions PCI Express* PCI Express*
Rx 2
I/O request
Packet Intel Ethernet Intel Data Direct I/O Intel Ethernet
read of
I/O data
Controller Technology Controller
(Intel DDIO)
23 Intel QPI - Intel QuickPath* Interconnect
24. 10GbE Completes the Job Faster
4X
Improvement
Performance comparison using best submitted/published 2-socket server results on the SPECfp*_rate_base2006 benchmark as of 6 March 2012.
Economies of scale realized with 10GbE
Software and workloads used in performance tests may have been optimized for performance only on Intel® microprocessors. Performance tests, such as
SYSmark* and MobileMark*, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors
may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including
the performance of that product when combined with other products. For more information go to http://www.intel.com/performance. Configuration:
Source: Intel internal measurements of average time for an I/O device read to local system memory under idle conditions comparing Intel® Xeon® processor
24 E5-2600 product family (230ns) vs. Intel Xeon processor 5500 series (340ns). See notes in backup for configuration details.
25. Platform and Software Optimizations
Integrated Up to four channels
PCI DDR3 1600 MHz
Express* memory
3.0
Up to 40
lanes Up to eight
per socket cores
Up to 20 MB
cache
• Up to 80% Performance Boost vs. Prior Generation1
– Intel® Advanced Vector Extensions (Intel® AVX) - Reduce
Compute Time
– Intel® Turbo Boost Technology — Increased Performance2
• Hadoop* Optimizations from Intel
– Built on Open Source Releases
– Custom Tuning for Data Types and Scaling Approaches
1 Performance comparison using best submitted/published 2-socket server results on the SPECfp*_rate_base2006 benchmark as of 6 March 2012.
25 2 Source: Intel internal measurements of average time for an I/O device read to local system memory under idle conditions comparing Intel® Xeon® processor E5-2600 product family (230ns) vs.. Intel Xeon
processor 5500 series (340ns). See notes in backup for configuration details
26. Intel® Intelligent Storage
Acceleration Library (Intel® ISA-L)
Algorithmic Library to address key
Storage market segment needs Normalized to Existing Open
Source Solutions
• Optimized for Intel® Architecture
4
• Enhances efficiency, data integrity,
security/encryption 3
Benefits of using Intel® ISA-L 2
• Allows maximum utilization of additional
1
cores
• Faster time to market (TTM)/less 0
resources than developing in-house
• Allows Intel to develop optimizations
using new architectural enhancements
that promote faster TTM AVX Multi-buffer Hashing Functions
(Baseline case is without Intel ISA-L)
Intel ISA-L enables OEMs to obtain more
performance from Intel® CPUs
26 Note: For more information go to http://www.intel.com/performance
27. A Fresh Look at Intel® Virtualization
Technology (Intel® VT)
Intel® Virtualization Technology
Traditional Server VMM Intel VT for Intel VT for
• Isolate development and IA-32 and Intel
64 (Intel VT-x)
Directed I/O
(Intel VT-d)
production environment HW support for HW support for
isolated execution isolated I/O
• Technology demonstrators
New Cloud Security
Model
• Isolation of workloads in
multi-tenant cloud VM1 VM2
• Memory monitoring for
malware detection
VMM
• Device Isolation for protection
against DMA attacks
Hardware Provides Stronger Isolation of VMs
27 Intel® Virtualization Technology for IA-32, Intel® 64 and Intel® Architecture (Intel® VT-x)
28. Intel® Trusted Execution Technology
(Intel® TXT)
Trusted Pools
Intel® TXT Control VMs based on
platform trust to better
• Enables isolation and tamper protect data
detection in boot process Trusted Launch
• Complements runtime Verified platform
integrity reduces
protections malware threat
• Hardware based trust provides
verification useful in Internet
compliance
• Trust status usable by security
and policy applications to
control workloads
Compliance
Hardware support for compliance
reporting enhances auditability of
cloud environment
Hardens and Helps Control the Platform
28
29. Data Protection with Intel® AES-NI
Data at Rest
Full disk encryption
software protects data
while saving to disk
Intel® AES-NI Data in Motion
• Special math functions built Secure transactions
used pervasively in
in the processor accelerate ecommerce, banking,
etc.
processing of crypto
algorithms like AES
• Includes 7 new instructions
Internet Intranet
• Makes enabled encryption
software faster and stronger
Data in Process
Most enterprise and cloud applications
offer encryption options to secure
information and protect confidentiality
Efficient Ways to Use Encryption for Data Protection
29 Intel® AES New Instructions (Intel® AES-NI)
30. Agenda
• What is Big Data?
• Why does Big Data matter?
• How can Intel®
Architecture help
• Summary
30
31. Big Data rEvolution
From To
Collecting Connecting
From To
Analyzing Predicting
From To
Structured Unstructured
31
32. Summary
Compute
Intel® Xeon® Processor • Big Data Phenomenon is
Real
• Analytics based on
Hadoop* will be the
norm
• Compute, Network &
Storage will converge
for Big Data Solutions
Network Storage
10GbE Network NVM, Tiered, JBOD
Intel® Architecture is foundational to finding “Your 5K”
32
34. Legal Disclaimer
• Intel® Hyper-Threading Technology (Intel® HT Technology) is available on select Intel® Core™ processors. Requires
an Intel® HT Technology-enabled system. Consult your PC manufacturer. Performance will vary depending on the
specific hardware and software used. For more information including details on which processors support Intel HT
Technology, visit http://www.intel.com/info/hyperthreading.
• Intel® Trusted Execution Technology (Intel® TXT): No computer system can provide absolute security under all
conditions. Intel® TXT requires a computer with Intel® Virtualization Technology, an Intel TXT enabled processor,
chipset, BIOS, Authenticated Code Modules and an Intel TXT compatible measured launched environment (MLE).
Intel TXT also requires the system to contain a TPM v1.s. For more information, visit
http://www.intel.com/technology/security
• Intel® Virtualization Technology (Intel® VT) requires a computer system with an enabled Intel® processor, BIOS,
and virtual machine monitor (VMM). Functionality, performance or other benefits will vary depending on hardware
and software configurations. Software applications may not be compatible with all operating systems. Consult your
PC manufacturer. For more information, visit http://www.intel.com/go/virtualization
• Intel® Turbo Boost Technology requires a system with Intel Turbo Boost Technology. Intel Turbo Boost Technology
and Intel Turbo Boost Technology 2.0 are only available on select Intel® processors. Consult your PC manufacturer.
Performance varies depending on hardware, software, and system configuration. For more information, visit
http://www.intel.com/go/turbo
• Intel® AES-NI requires a computer system with an AES-NI enabled processor, as well as non-Intel software to
execute the instructions in the correct sequence. AES-NI is available on select Intel® processors. For availability,
consult your reseller or system manufacturer. For more information, see
Intel® Advanced Encryption Standard Instructions (AES-NI)
34
35. Risk Factors
The above statements and any others in this document that refer to plans and expectations for the second quarter, the year and the
future are forward-looking statements that involve a number of risks and uncertainties. Words such as “anticipates,” “expects,”
“intends,” “plans,” “believes,” “seeks,” “estimates,” “may,” “will,” “should” and their variations identify forward-looking statements.
Statements that refer to or are based on projections, uncertain events or assumptions also identify forward-looking statements.
Many factors could affect Intel’s actual results, and variances from Intel’s current expectations regarding such factors could cause
actual results to differ materially from those expressed in these forward-looking statements. Intel presently considers the following
to be the important factors that could cause actual results to differ materially from the company’s expectations. Demand could be
different from Intel's expectations due to factors including changes in business and economic conditions, including supply constraints
and other disruptions affecting customers; customer acceptance of Intel’s and competitors’ products; changes in customer order
patterns including order cancellations; and changes in the level of inventory at customers. Uncertainty in global economic and
financial conditions poses a risk that consumers and businesses may defer purchases in response to negative financial events, which
could negatively affect product demand and other related matters. Intel operates in intensely competitive industries that are
characterized by a high percentage of costs that are fixed or difficult to reduce in the short term and product demand that is highly
variable and difficult to forecast. Revenue and the gross margin percentage are affected by the timing of Intel product introductions
and the demand for and market acceptance of Intel's products; actions taken by Intel's competitors, including product offerings and
introductions, marketing programs and pricing pressures and Intel’s response to such actions; and Intel’s ability to respond quickly
to technological developments and to incorporate new features into its products. Intel is in the process of transitioning to its next
generation of products on 22nm process technology, and there could be execution and timing issues associated with these changes,
including products defects and errata and lower than anticipated manufacturing yields. The gross margin percentage could vary
significantly from expectations based on capacity utilization; variations in inventory valuation, including variations related to the
timing of qualifying products for sale; changes in revenue levels; segment product mix; the timing and execution of the
manufacturing ramp and associated costs; start-up costs; excess or obsolete inventory; changes in unit costs; defects or disruptions
in the supply of materials or resources; product manufacturing quality/yields; and impairments of long-lived assets, including
manufacturing, assembly/test and intangible assets. The majority of Intel’s non-marketable equity investment portfolio balance is
concentrated in companies in the flash memory market segment, and declines in this market segment or changes in management’s
plans with respect to Intel’s investments in this market segment could result in significant impairment charges, impacting
restructuring charges as well as gains/losses on equity investments and interest and other. Intel's results could be affected by
adverse economic, social, political and physical/infrastructure conditions in countries where Intel, its customers or its suppliers
operate, including military conflict and other security risks, natural disasters, infrastructure disruptions, health concerns and
fluctuations in currency exchange rates. Expenses, particularly certain marketing and compensation expenses, as well as
restructuring and asset impairment charges, vary depending on the level of demand for Intel's products and the level of revenue and
profits. Intel’s results could be affected by the timing of closing of acquisitions and divestitures. Intel's results could be affected by
adverse effects associated with product defects and errata (deviations from published specifications), and by litigation or regulatory
matters involving intellectual property, stockholder, consumer, antitrust, disclosure and other issues, such as the litigation and
regulatory matters described in Intel's SEC reports. An unfavorable ruling could include monetary damages or an injunction
prohibiting Intel from manufacturing or selling one or more products, precluding particular business practices, impacting Intel’s
ability to design its products, or requiring other remedies such as compulsory licensing of intellectual property. A detailed discussion
of these and other factors that could affect Intel’s results is included in Intel’s SEC filings, including the company’s most recent Form
10-Q, Form 10-K and earnings release.
Rev. 5/4/12
35