27
Akanksha Balani Regional Lead, Developer Products Intel® Software

Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

  • Upload
    others

  • View
    5

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Akanksha BalaniRegional Lead, Developer Products

Intel® Software

Page 2: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Agenda

- Turbocharge your Software using Intel Software tools

- Introduction to Intel Parallel Studio XE & ML/DL capabilities

- Value Intel Software can bring to the table

- Intel's leading Platform for AI Capabilities for Finance

- Customers across the world benefitting from AI through SW

- Way Forward

Page 3: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Intel® Xeon®

processor

64-bit

Intel® Xeon®

processor 5100

series

Intel® Xeon®

processor

5500 series

Intel® Xeon®

processor

5600 series

Intel® Xeon®

processor

5600 v2

series

Intel® Xeon®

processor

5600 v3

series

BDW

Up to Core(s) 1 2 4 6 12 18 28

Up to Threads 2 2 8 12 24 36 44

SIMD Width 128 128 128 128 256 256 256

Vector ISAIntel® SSE3

Intel® SSE3

Intel® SSE4.2

Intel® AVX

Intel® AVX

Intel® AVX2

Intel® AV512

More cores More Threads Wider vectors

Compute Evolution Enhances the need for Code Optimization

3

Page 4: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice4

Volatile markets and complex regulations are

placing ever-higher demands on trading, risk

management, and other platforms. The Intel®

Xeon® processor product family is helping vendors

answer the challenge. By churning through more

computations in less time, these processors can help

organizations make smarter decisions in high-

volume scenarios where every microsecond matters.

Page 5: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Why Use Software Tools? - Turbocharge your software

Approach Definition Tools of the Trade

Thread Scaling

Increase concurrencyacross coherent shared memory

OpenMP*, Intel® Threading Building Blocks (TBB)

Vector Scaling

Use wide vector instructions

Vector loops, vector functions, array notation

Cache Blocking

Use algorithms to reduce memory bandwidth pressure and improve cache reuse

Blocking algorithms

Fabric Scaling

Distribute work across connected systems

MPI

Data Layout

Optimize data layout forunconstrained performance

AoSSoA, directives for alignment

X4

Y4

Z4

X3

Y3

Z3

X2

Y2

Z2

X1

Y1

Z1

0X8

Y8

Z8

X7

Y7

Z7

X6

Y6

Z6

X5

Y5

Z5

X12

Y12

Z12

X11

Y11

Z11

X10

Y10

Z10

X9

Y9

Z9

X16

Y16

Z16

X15

Y15

Z15

X14

Y14

Z14

X13

Y13

Z13

512

6

Page 6: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

How your Business can Benefit from More Performance

6

Use the full power, get more performance from Intel hardware

Speed applications and workload processing

Increase efficiency and developer productivity

Solve business challenges, fuel innovation

Scale forward, drive compatibility and interoperability

Intel® Parallel Studio XE

Boosts performance on today’s and future Intel® platforms

Simplifies creating high performance, scalable, reliable parallel code with less effort and performance on data analytics/ML/DL.

Page 7: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization NoticeCopyright© 2017, Intel Corporation. All rights reserved.

Which intel software suite?

7

Build highly optimized media infrastructure, solutions, & applications

Fast, Dense, High Quality Transcoding

Improve performance, scalability, & reliability for parallel applications

Technical & Enterprise compute, HPC, AI

Take advantage of deep system-wide insight & analysis for system & embedded apps

Manuf., Retail, Drones, Robots…

Smart Cities, Auto. Driving, Gaming…

Create solutions using Computer Vision, Deep Learning, Graphics, Libraries, Media, OpenCL™, & more

Optimization Tools & SDKs

Edge to DC to Cloud

Page 8: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Cluster EditionProfessional EditionComposer Edition

Intel® Software tools - Create Faster Code…Faster

8

More Power for Your Code - software.intel.com/intel-parallel-studio-xe

Intel® VTune™ AmplifierPerformance Profiler

ANALYZEAnalysis Tools

Intel® AdvisorVectorization Optimization

& Thread Prototyping

Intel® InspectorMemory & Thread Debugger

SCALECluster Tools

Intel® Trace Analyzer & CollectorMPI Tuning & Analysis

Intel® MPI LibraryMessage Passing Interface Library

Intel® Cluster CheckerCluster Diagnostic Expert System

Operating System: Windows*, Linux*, MacOS1*

Intel® Architecture Platforms

BUILDCompilers & Libraries

C / C++ CompilerOptimizing Compiler

Intel® Distribution for Python*High Performance Scripting

Intel® MKLFast Math Kernel Library

Intel® IPPImage, Signal & Data Processing

Intel® TBBC++ Threading Library

Intel® DAALData Analytics Library

Fortran CompilerOptimizing Compiler

Page 9: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Intel® Parallel Studio XE Bottom Line – ROI Analysis

ROI Analysis DetailsIntel® Parallel Studio XE

for Windows*

Intel® Parallel Studio XE

for Linux*

Assuming Annual Developer Burden Costs6:$125,000 $125,000

Calculated Developer Weekly Salary: $2,604 $2,604

Product Price (1st year includes maintenance):$2,299 $2,299

Yearly maintenance: $759 $899

Developer weeks time savings: 6.2 weeks 6.2 weeks

Developer cost savings 6.2 weeks x 2,604 = $16,275

Plus $5,000 annual savings by catching defects early$21,275 $21,275

Payback period: .5 month .5 month

ROI first year 1120% 946%

Annual Savings/developer:$21,275

First Year ROI: 1120%

Payback period:.5 month

Intel® Parallel Studio XEMaximizing ROI

• Increased Developer productivity• Higher application quality and security• Optimized performance and scalability• Faster ramp to modernizing code.

Link to White Paper: ROI from Optimizing SW Performance with Intel® Parallel Studio XE

Page 10: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Intel® Tools Competitive Advantages

10

Intel® Compiler Visual C++ GCC*

Portland Group* (NVIDIA)

Performance leadership on systems with Intel & compatible AMD CPUs (SPEC)

Y N N N

Enhanced vectorization support Y Y, limited N Y, substantial extra effort

(CUDA)

Support for various architectures (Xeon to Atom)

Y N Y N

Commercial customer support Y Y N Y, extra $$

Page 11: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

4X 8X 35%Kyoto University

the Walker MolecularDynamics lab 3X

1.4X 4X

Energy

EDA

Science & Research

Manufacturing

Government

Computer Software

IT

Healthcare

Digital Media

Telecommunications

Intel® Parallel Studio XE: High Performance, Scalable Software across Multiple Industries

View case studies details online.

10X

11X

25X

2.5X 1.25X 1.3X

5X 2X

20X

2.5X

Software & workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark & MobileMark, are measured using specific computer systems, components, software, operations & functions. Any change to any of those factors may cause the results to vary. You should consult other information & 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.

11

Optimization Notice: Intel's compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the avai lability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice.

Page 12: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice12

FSI Community Enabled

Page 13: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Financial ServicesType of application to look out for

– Risk modeling, quantitative analysis or Quants, Risk Mitigation

– Options trading and derivatives, market feeds, Algorithmic Trading

– Monte Carlo simulation, binomial, trinomial option pricing

– Cryptocurrency

– Fraud Detection

– Research

– Personal Finance

Customer Needs software tools address

– New instruction support, load balancing

– Binomial pricing market Risk Analysis

– increasing numpy/scipy performance

– ML/DL capabilities

Page 14: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

libraries Intel® Math Kernel Library (MKL, MKL-DNN)

platforms

Frameworks

Intel® Data Analytics Acceleration Library

(DAAL)

hardwareMemory & Storage NetworkingCompute

Intel® Python Distribution

Mllib BigDL

Intel® Nervana™ Graph*

experiences

Intel® Nervana™ Cloud & System

Intel® Nervana™ portfolio

Intel® Nervana™ Deep Learning Studio

Intel® Computer Vision SDK

Movidius™Technology

*FutureOther names and brands may be claimed as the property of others.

*

Page 15: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice15

Using Intel® MKL for Financial Mathematics

Page 16: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Intel® Data Analytics Acceleration Library

An industry leading, open source, C++/Java/Python library for machine learning and deep learning optimized for Intel® Architectures.

(De-)CompressionPCAStatistical momentsVariance matrixQR, SVD, CholeskyApriori

Linear regressionNaïve BayesSVMClassifier boosting

KmeansEM GMM

Collaborative filtering

Neural Networks

Pre-processing Transformation Analysis Modeling Decision Making

Sci

en

tifi

c/E

ng

ine

eri

ng

We

b/S

oci

al

Bu

sin

ess

Validation

Page 17: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

What’s Inside Intel® Integrated Performance PrimitivesHigh Performance , Easy-to-Use & Production Ready APIs

1 Available only in Intel® Parallel Studio Composer Edition.

Operating System: Windows*, Linux*, MacOS1*

Intel® Architecture Platforms

Page 18: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

github.com/intel-analytics/BigDL

Open Sourced Deep Learning Library for Apache Spark*

Make Deep learning more Accessible to Big data users and data scientists.

Feature Parity with popular DL frameworks like Caffe, Torch, Tensorflow etc.

Easy Customer and Developer Experience Run Deep learning Applications as Standard

Spark programs;

Run on top of existing Spark/Hadoop clusters (No Cluster change)

High Performance powered by Intel MKL and Multi-threaded programming.

Efficient Scale out leveraging Spark architecture.

Spark Core

SQL SparkRStream-

ingMLlib GraphX

ML Pipeline

DataFrame

BigDL

For developers looking to run deep learning on Hadoop/Spark due to familiarity or analytics use

BigdlBringing Deep Learning to Big Data

Page 19: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

software.intel.com/intel-distribution-for-python

Easy, Out-of-the-box Access to High Performance Python

Prebuilt, optimized for numerical computing, data analytics, HPC

Drop in replacement for your existing Python (no code changes required)

Drive Performance with Multiple Optimization

Techniques

Accelerated NumPy/SciPy/Scikit-Learn with Intel® MKL

Data analytics with pyDAAL, enhanced thread scheduling with TBB, Jupyter* Notebook interface, Numba, Cython

Scale easily with optimized MPI4Py and Jupyter notebooks

Faster Access to Latest Optimizations for Intel

Architecture

Distribution and individual optimized packages available through conda and Anaconda Cloud

Optimizations upstreamed back to main Python trunk

For developers using the most popular and fastest growing programming language for AI

Intel distribution for pythonAdvancing Python Performance Closer to Native Speeds

Page 20: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Distribution Details

Open Source

Apache 2.0 License

Common DNN APIs across all Intel hardware.

Rapid release cycles, iterated with the DL community, to best support industry framework integration.

Highly vectorized & threaded for maximal performance, based on the popular Intel® MKL library.

For developers of deep learning frameworks featuring optimized performance on Intel hardware

github.com/01org/mkl-dnn

Direct 2D Convolution

Rectified linear unit neuron activation

(ReLU)

Maximum pooling

Inner productLocal response normalization

(LRN)

Intel® MKL-dnnMath Kernel Library for Deep Neural Networks

Examples:

Page 21: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Page 22: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice22

TATA Consultancy Services Ltd* - BigData Analytics for Security Trading

Page 23: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice23

BigDL Use case: Transaction Fraud DetectionFraud detection system based on BigDL to increase fraudulent transaction detection coverage and accuracy for a leading credit card provider

• Historical data is stored on Hive• Data preprocessing with SparkSQL• Spark ML pipeline for complex feature

engineering• Use multiple BigDL CNN models• Use Sample+Bagging to solve

unbalance problem• Grid search for hyper parameter tuning Powered by BigDL

Page 24: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice24

Accelerate Intelligent Solutions with Ia Machine Learning Platform

China UnionPay is seeing a significant growth in its transaction volumes. This brings

opportunities but also a corresponding increase in fraud threats and other risks

China UnionPay implemented a neural-network risk-control system based on robust

Apache Spark* computing clusters and it's deep learning library – BigDL- accelerated

with Intel Math Kernel Library, this uses machine learning to deliver greater consistency

and efficiency while significantly reducing man-hours for user development, deployment

and maintenance

Challenge

Solution

Page 25: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice25

Intel® Nervana™ AI Academy

Intel® Nervana™ DevCloud

Intel® Nervana™ Cloud

Intel Software Tools

Intel Optimized Frameworks

Building Blocks for Intel on AI –

Be Part of the Innovation

Today!

Page 26: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster

Copyright © 2017, Intel Corporation. All rights reserved. *Other names and brands may be claimed as the property of others.

Optimization Notice

Legal Disclaimer & Optimization Notice

INFORMATION IN THIS DOCUMENT IS PROVIDED “AS IS”. NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. INTEL ASSUMES NO LIABILITY WHATSOEVER AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO THIS INFORMATION INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT.

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.

Copyright © 2015, Intel Corporation. All rights reserved. Intel, Pentium, Xeon, Xeon Phi, Core, VTune, Cilk, and the Intel logo are trademarks of Intel Corporation in the U.S. and other countries.

Optimization Notice

Intel’s compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice.

Notice revision #20110804

26

Page 27: Akanksha Balani Regional Lead, Developer Products Intel ...€¦ · TBB, Jupyter* Notebook interface, Numba, Cython Scale easily with optimized MPI4Py and Jupyter notebooks Faster