Accelerating ML Training with AI

Tools from Intel®

24th January, 2024
 About our Speakers
Janakiram MSV– Principal Analyst, Janakiram & Associates
Janakiram MSV (Jani) is a practicing architect, research analyst, and advisor to Silicon Valley startups.
He focuses on the convergence of modern infrastructure powered by Kubernetes and machine
intelligence driven by IoT, Edge, and AI. Before becoming an entrepreneur, he spent over a decade
working as a product manager and technology evangelist at Microsoft Corporation and Amazon Web
Services. Janakiram regularly writes for Forbes, InfoWorld, and The New Stack, covering the latest from
the technology industry. He is a keynote speaker for internal sales conferences, product launches, and
user conferences hosted by technology companies of all sizes..

Susan Kahler - AIML Technical Product Marketing Manager

Susan is a Technical Product Marketing Manager for AIML at Intel. She has her Ph.D. in Human Factors
and Ergonomics, having used analytics to quantify and compare mental models of how humans learn
complex operations. Throughout her well-rounded career, she has held roles in user-centered design,
product management, customer insights, consulting, and operational risk.

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 2
 Resize and move windows around

Using the Console

Biographies

Resources
Presentation

Speaker video
Live demos

Hover over the video box to access closed captioning Quick links Questions and Answers

Department
Webinar or Event Name
Title Here Intel Confidential 3
 Agenda
• Introduction to AI Tools from Intel
• Getting Started with AI Tools from Intel
• Accelerate Pandas DataFrame Processing
• Comparing Pandas vs. Intel Modin
• Accelerate scikit-learn for Machine Learning
• Comparing Intel® Extension for Scikit-learn vs. Sklearn

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 4
 Introduction to AI Tools from Intel®
• AI Tools from Intel
• Provides Python-based tools and frameworks for data scientists, AI developers, and
researchers to enhance data science and analytics pipelines on Intel® architecture, utilizing
oneAPI libraries for computational efficiency.
• Low-Level Compute Optimizations
• Built on oneAPI libraries, these tools offer optimized performance across various stages of
AI development, from preprocessing to machine learning.
• Enhanced AI Workflow
• Enables training on Intel® CPUs and GPUs, integrates fast inference with Intel®-optimized
frameworks for TensorFlow and PyTorch, and includes pretrained models and model
optimization tools for efficient AI development.
• Accelerated Data Preprocessing and Machine Learning
• Offers drop-in acceleration for compute-intensive Python packages, including Modin, scikit-
learn, and XGBoost, streamlining data preprocessing and machine learning workflows.
• Seamless Software Integration
• Provides direct access to analytics and AI optimizations from Intel, ensuring interoperability
and seamless integration of software components in AI projects.

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 5
 Getting Started with AI Tools from Intel®
• Download the AI Tools
• Access the tools on the Intel website to begin working with optimized deep learning
frameworks and high-performing Python libraries.
• Develop in the Cloud
• Utilize Intel's cloud services to build and optimize multiarchitecture applications with
the latest Intel-optimized oneAPI and AI tools.
• Choose Your AI Platform
• Select the appropriate Intel CPU and GPU hardware for your AI projects, ensuring
compatibility and optimized performance.
• Utilize Optimized Frameworks
• Leverage Intel-optimized frameworks such as TensorFlow and PyTorch, along with
pretrained models and model optimization tools for efficient development.
• Explore Documentation and Code Samples
• Make use of the available documentation and code samples on Intel's website to guide
your development process and learn best practices.

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 6
 Accelerate Pandas DataFrame Processing
• Modin is a drop-in replacement for pandas, enabling data
scientists to scale to distributed DataFrame processing without
having to change API code.
• Intel® Distribution of Modin adds optimizations to further
accelerate processing on Intel hardware.
• Using this library, Developers can:
• Process terabytes of data on a single workstation
• Scale from a single workstation to the cloud using the same code
• Focus more on data analysis and less on learning new APIs

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 7
 Accelerate Pandas DataFrame Processing

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 8
 Accelerate Pandas DataFrame Processing

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 9
 Scale Pandas Workloads with a Single Code-
Line Change
Demo

Department
Webinar or Event Name
Title Here Intel Confidential 10
 Accelerate Scikit-learn for Machine Learning
• Intel® Extension for Scikit-learn seamlessly speeds up scikit-learn
applications for Intel CPUs and GPUs across single- and multi-
node configurations.
• The extension is part of the AI Tools that provide flexibility to use
machine learning tools with existing AI packages.
• Using scikit-learn with this extension, developers can:
• Speed up training and inference by up to 100x with the equivalent
mathematical accuracy.
• Continue to use the open source scikit-learn API.
• Enable and disable the extension with a couple lines of code or at the
command line.

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 11
 Accelerate Scikit-learn for Machine Learning

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 12
 Accelerate Scikit-learn for Machine Learning

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 13
 Scale ML Training and Inference a Single
Code-Line Change
Demo

Department
Webinar or Event Name
Title Here Intel Confidential 14
 Find more workshops, webinars, and events here!

• One API Events and Training calendar

-https://software.seek.intel.com/oneapi-training-calendar#gs.8qflef

• Intel Software Developer Zone events

- calendar#gs.8qf9ffhttps://www.intel.com/content/www/us/en/events/developer/overview.html

• Upcoming Webinars
-https://software.seek.intel.com/techdecoded-webinars

Optimization Notice
Optimization Notice
Copyright © 2022, Intel Corporation. All rights reserved. 15
*Other names and brands may be claimed as the property of others.
The Intel® Academic Program for oneAPI
Fostering the advancement of innovation in the academic ecosystem

Teach. Lead.
Educator Program oneAPI Centers of
Excellence
Collaborating with
professors in teaching Learn. Accelerating
world-changing technology the open standards-based
Intel® Student programming model by enabling
to tomorrow’s developers
Ambassador for oneAPI widely used codes
Investing in students to
become leaders in their
academic developer
community

The Intel® Academic Program for oneAPI collaborates with academia worldwide
to enable the next generation of developers, scientists and engineers to
advance accelerated computing.

LEARN MORE
 Additional Academic Offerings

Teaching Kits Intel® DevMesh DevSummits Intel® Developer Learning Paths

Cloud
▪ Resources include videos, ▪ Sharing Projects, Code ▪ Flagship developer ▪ Curated set of training
presentations, assignments, Repos, Videos and events held around ▪ oneAPI and OpenVINO modules that help you
and quizzes for Educators Blogs globe virtually and in- learn specific tools and
▪ Access to a wide range
person technologies.
▪ Topics include: ▪ AI, IOT, oneAPI and of Intel® architectures
Visual Computing ▪ 80% presentations
▪ SYCL Programming for
▪ 100+ quality oneAPI from academia and
Accelerated Computing
and AI projects industry
▪ Introduction to Machine
Learning
▪ Intel® Distribution of
OpenVINO™ Toolkit
Intel® Software Innovator Program
Get the technology tools and resources to share achievements in a global peer community

Access to Intel Hardware and

Intel Expertise

Early Code, Testing & Research

Spotlight Opportunities for

Speakerships
APPLY TODAY

Meet oneAPI community of highly

regarded researchers, software
developers, and students
 Intel® Certified Instructor Program for oneAPI
Endorsing qualified third-party developers to teach oneAPI content

Current topics:
Program Benefits C++ SYCL
Machine Qualifying Criteria
Learning for
Essentials
oneAPI
• Experience in High Performance
• Extended Intel® Developer Cloud access
oneAPI
Computing, Data Science, Computer
• Intel® Certified Instructor badge and OpenMP*
Rendering Vision, Machine Learning or Artificial
Offload Basics
certificate (per topic) Toolkit Intelligence
• Featured profile on oneAPI Expert page • Demonstration of technical mastery in
and other Intel communication channels CUDA the topic of instructor certification
Migration
• Editable PowerPoint presentations with • Instructor ability showcased through a
speaker notes and additional submitted instruction sample
instructional resources Upcoming topics: • Support of Intel’s cross architecture
• Ongoing training and product updates by vision using oneAPI
Deep
Intel experts Learning for
oneAPI

GPU
Optimalization

*Annual certification renewal is required. The renewal process involves continuing education and a record of instructor activity
APPLY TODAY
 Notices & Disclaimers

Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.
Performance results are based on testing as of dates shown in configurations and may not reflect all publicly available updates. See backup for
configuration details.
No product or component can be absolutely secure.
Your costs and results may vary.
Intel technologies may require enabled hardware, software or service activation.
Intel does not control or audit third-party data. You should consult other sources to evaluate accuracy.
© Intel Corporation. Intel, the Intel logo, Xeon, Core, VTune, OpenVINO, and other Intel marks are trademarks of Intel Corporation or its subsidiaries.
Other names and brands may be claimed as the property of others.

Department
Accelerating orTraining
ML Eventwith
Name
AI Tools from Intel® Intel Confidential 20