Hi [[ session.user.profile.firstName ]]

Bringing PyTorch Models to Cortex-M Processors

This talk will take you step-by-step through how to develop a deep learning model with PyTorch. Rohit Sharma from AI Tech Systems (AITS), a leading firmware, software and services provider for low power IoT, Endpoint and tinyML devices, will demonstrate how PyTorch can be used on Arm Cortex-M based systems. Get an in-depth look at how to bring on-device AI applications to a multitude of verticals, including Industrial IoT, smart space and much more as AITS showcase a live demo using Arduino and STM32F07 boards.
Recorded Jun 28 2021 59 mins
Your place is confirmed,
we'll send you email reminders
Presented by
Rohit Sharma, AITS
Presentation preview: Bringing PyTorch Models to Cortex-M Processors
  • Channel
  • Channel profile
  • Disrupting Cloud Data Centers with Arm Neoverse Aug 12 2021 10:00 am UTC 60 mins
    Eddie Ramirez, Senior Director Marketing, Infrastructure Business Unit, Arm
    Technology infrastructure is quickly evolving, to keep in step with the rapid pace of digital transformation.

    The expansion and adoption of IoT, 5G, Intelligent Edge and AI/ML adds further momentum to high performance computing and secure network infrastructure.

    At the heart of this digital transformation is the data center infrastructure needed to deliver secure, scalable leading-edge performance, with efficiencies to cut cost and create competitive advantages.

    Arm's Neoverse Platform is the foundation for the next generation Cloud-to-Edge Infrastructure, delivering high-performance, secure and scalable computing solutions along with a robust hardware and software ecosystem.

    Arm is rapidly expanding its footprint in cloud infrastructures, evidenced by having cloud service providers and hyperscalers like AWS, Oracle Cloud (OCI), Microsoft Azure, Tencent, ByteDance and several others as customers of Arm Neoverse server solutions. A broad array of workloads from cloud-native applications, SaaS, EDA, HPC, Big Data, Analytics and AI/ML, are optimized and deployed on Arm’s infrastructure technologies.

    This webinar will highlight :

    - Arm’s Neoverse Platform, a scalable, secure and efficient infrastructure technology solution
    - Comprehensive hardware and software ecosystem, that enable and optimize your application development and deployment on Arm based infrastructure.
    - Success stories with large hyperscale data centers.

    Are you ready to reap the benefits of high performance, cost-efficient and secure infrastructure in your cloud and on-premise data center?

    Join us to learn how Arm's Neoverse platform and ecosystem partners can help you foster innovation and growth with successful deployment in the Hyperscale and Enterprise cloud data centers.

    For more information on how Arm processes your personal data, please visit the Arm Privacy Policy available here: https://www.arm.com/company/policies/privacy
  • Next Gen Authentication Without the Cloud Jul 30 2021 12:00 am UTC 60 mins
    Peter Padd, Malcolm Layton, David Chung, Dr. Henry Weld from Fortifyedge & Mary Bennion, Arm
    Manufacturers of mobile devices can focus on product development with no worries around security, thanks to a pioneering company who have created a way of authenticating users without the need for a password.

    Learn more in this webinar featuring Arm AI Partner Program member Fortifyedge.

    They will provide a demo of their Zero Trust Authentication Software which utilizes neural networks to continually authenticate users and connected IoT devices, without the need for the cloud.

    In this webinar we will cover:
    - How Fortifyedge has developed Zero Trust Authentication Software to provide OEMs developing products with a password-free authentication process
    - How the technology utilizes next generation behavioral biometrics on Edge devices such as smartphones, wearables and smart IoT devices
    - What happens to the password. With security taken care of, manufactures can focus on delivering next generation smart products.

    For more information on how Arm processes your personal data, please visit the Arm Privacy Policy available here: https://www.arm.com/company/policies/privacy
  • Future-Proof Vision Systems for ADAS Jul 28 2021 9:00 am UTC 45 mins
    Radhesh Bhat, PathPartner, Dave Tokic, Algolux, Graeme Voller, Arm & Priyanka Viswanathan, Arm
    In this panel discussion you will hear how Arm’s safety-capable Mali Image Signal Processors (ISPs) work together with various IQ tuning and software tools and services to develop a future-proof vision system for ADAS.

    • ADAS vision system workloads
    • Safety for computer vision applications
    • Effects of compromised vision performance

    This webinar is part of a 3 part series. Links to the other webinars can be found in the attachments section.

    For more information on how Arm processes your personal data, please visit the Arm Privacy Policy available here: https://www.arm.com/company/policies/privacy
  • Functional Safety Development Solutions for Arm Cortex-M Recorded: Jul 21 2021 26 mins
    Maxim Damm, Siemens, Loic Chossat, STMicroeletronics, Dr Daniel Kästner, AbsInt, Christopher Seidl & Vladimir Marchenko, Arm
    In this session, you’ll learn how pre-qualified software components and libraries help you to get to a certified product in less time when developing functional safety applications on Arm-based microcontrollers. We will also touch base on tool qualification, requirements capturing, and static code analysis and how all these methods interact with each other.

    • Software test libraries
    • Requirements capturing
    • Static code analysis

    This webinar is part of a 3 part series. Links to the other webinars can be found in the attachments section.

    For more information on how Arm processes your personal data, please visit the Arm Privacy Policy available here: https://www.arm.com/company/policies/privacy
  • Achieving Safety-Critical Rendering with Mali-G78AE, CoreAVI and DiSTI Recorded: Jul 14 2021 42 mins
    Christopher Giordano, DiSTI, Mustafa Ali, Arm, Neil Stroud, CoreAVI & Antonio Priore, Arm
    In this panel discussion you will hear about how a safety-critical graphics rendering solution for digital cockpits can be built with the Arm Mali-G78AE GPU and the safety critical driver stack developed through the Arm-CoreAVI partnership.

    We will also be talking about how automotive OEMs can use this solution to deploy functionally safe graphics using DiSTI GL Studio.

    • The importance of safety-critical rendering in modern and future digital cockpits
    • The capability needed in the supply chain to meet the demands for rendering rich, high resolution 3D graphics safely by OEMs
    • How the Mali-G78AE GPU along with DiSTI’s HMI tooling help to create safety-critical graphics rendering solution for digital cockpits
    • What is importance of mixed-critical design and how this is supported

    This webinar is part of a 3 part series. Links to the other webinars can be found in the attachments section.

    For more information on how Arm processes your personal data, please visit the Arm Privacy Policy available here: https://www.arm.com/company/policies/privacy
  • Process Isolation in Safety-Critical Systems with Arm FuSa RTS Recorded: Jul 7 2021 58 mins
    Vladimir Marchenko, Product Manager MDK Core and Middleware, Arm
    Arm Run-Time System for functional safety (FuSa RTS) provides process isolation capabilities that allow integration of software components with different safety integrity levels (SIL) on a single-core MCU platform.

    This webinar explains process isolation features in details and highlights the benefits this brings to developers. The functionality is demonstrated with a complex example that among others integrates TCP/IP stack and Software-Test-Libraries (STL).

    For more information on how Arm processes your personal data, please visit the Arm Privacy Policy available here: https://www.arm.com/company/policies/privacy
  • Meet Arm Keil Studio - The New Cloud-Based IDE to Simplify IoT Development Recorded: Jun 30 2021 50 mins
    Christopher Seidl, Senior Product Manager Embedded Tools, Arm & Joe Alderson, Director of Embedded Tools and UX, Arm
    The Internet of Things (IoT) is transitioning into a new, complex world driven by increased compute, connectivity, and security capabilities, the rapid evolution of machine learning processing at the endpoint, 5G deployments, and evolving IoT platforms from major cloud service providers.

    Successful IoT implementations require strong industry collaboration and a robust ecosystem of support and tools. Arm's next-generation tool suite, Keil Studio will provide a step-change in productivity for IoT, ML, and embedded MCU developers, simplifying development and helping developers move quickly from production to prototype.

    During this webinar, Senior Product Manager, Christopher Seidl and Director of Embedded Tools and UX, Joe Alderson will introduce Keil Studio Cloud, the first component of the next generation Keil development tool suite. Christopher will share the tool's features and walk through some example projects covering:

    - How to get your code running on a device in minutes with hassle-free installation and example projects
    - Direct Git integration for enabling distributed teams, collaborative development, and modern CI workflows
    - Code editing with Intellisense and web-enabled debugging

    Combined with the renowned Keil MDK, this new offering will provide the best-in-class IoT, ML, and embedded development environment even for the most demanding real-time and FuSa projects.

    For more information on how Arm processes your personal data, please visit the Arm Privacy Policy available here: https://www.arm.com/company/policies/privacy
  • Introducing Netspresso – For Really Fast Inference on Cortex-A Devices Recorded: Jun 28 2021 53 mins
    Tae-Ho Kim, CTO & Co-Founder, Nota
    This talk takes you step-by-step through how Nota was able to compress different deep learning models on ARM devices including ResNet18 on Cortex-A72. By leveraging NetsPressso, a model compression platform built by Nota, you can optimize deep learning models in a range of environments and use cases with no learning curve.

    You will also get a detailed look at how this technology is a key component of commercialized lightweight deep learning models for facial recognition, intelligent transportation system, etc.

    If you need to deploy deep learning models on low-end devices, this talk is for you.
  • Automate tinyML Development and Deployment Recorded: Jun 28 2021 55 mins
    Tina Shyuan, Director of Product Marketing, Qeexo
    What’s the smallest machine learning model you’ve built? Qeexo’s is so small, it can even run on an Arm Cortex M0+! Experience how easy it is to automate "tinyML" machine learning development for sensor modules – without having to write a single line of code!

    In this workshop-style Tech Talk, we will:
    - Give an overview of the benefits and challenges of running machine learning at the edge.
    - Walk attendees through the installation and setup of Qeexo AutoML. Please come with a Windows or Mac laptop with the Google Chrome browser installed.
    - Demo the simple data collection and visualization with our Qeexo AutoML interface for attendees to follow along to build their own classifiers.
    - Build multiple machine learning models and deploy them to an Arm Cortex-M4-powered sensor module for live-testing with just a few clicks.
  • Optimised C-Code Generation for Ultra-Efficient tinyML Applications Recorded: Jun 28 2021 58 mins
    Johan Malm, ML Developer, Imagimob
    Writing C code for embedded systems is known to be slow and error prone. The solution? Imagimob Studio. It can turn a .h5-file containing a machine learning model into C code with only a click of a button. Optimizations for new AI accelerators and new network layers are easily added to Imagimob Studio. The C code can run on any platform since no external runtime libraries are needed. This talk uses different applications from real customers as examples, such as audio applications, condition monitoring and gesture control.
  • Bringing PyTorch Models to Cortex-M Processors Recorded: Jun 28 2021 59 mins
    Rohit Sharma, AITS
    This talk will take you step-by-step through how to develop a deep learning model with PyTorch. Rohit Sharma from AI Tech Systems (AITS), a leading firmware, software and services provider for low power IoT, Endpoint and tinyML devices, will demonstrate how PyTorch can be used on Arm Cortex-M based systems. Get an in-depth look at how to bring on-device AI applications to a multitude of verticals, including Industrial IoT, smart space and much more as AITS showcase a live demo using Arduino and STM32F07 boards.
  • Optimizing Android Graphics-Episode 3: Performance Analysis on Arm Mobile Studio Recorded: May 25 2021 81 mins
    Julie Gaskin, Staff Developer Evangelist, Arm
    Join us live for all three sessions to put your questions to the experts.

    This training series for game developers covers the latest techniques for optimizing mobile game performance. It will give you an in-depth training from fundamentals of underlying GPU hardware to API best practices. Hear first-hand advice from optimization experts and architects of the most prevalent graphics processors across Android devices today. By attending all three webinars you will learn tips for creating more efficient graphics, avoiding game lag and excessive battery usage, and how to discover graphics issues quickly using the free tools available in Arm Mobile Studio. The later session will introduce the modern workflows and tools that help you put in practise the principles that you learnt in the earlier sessions.
    At the end of each webinar, our experts will be available for an extended Q&A session where you can ask questions related to your own specific problems.

    Our third webinar introduces Arm’s free-to-use performance analysis tools, revealing faster techniques for identifying performance problems on off-the-shelf Android devices. We will demonstrate how Arm Mobile Studio lets you monitor FPS and see how your game is using the device’s CPU and GPU resources. You will also learn how to debug your graphics API calls and shader programs to find more ways to optimize your content for mobile.

    Content Covered:
    • Introducing Arm Mobile Studio
    • Identifying game bottlenecks quickly with Performance Advisor
    • Interpreting interactive charts in Streamline profiler
    • How to trace API calls using Graphics Analyzer
    • Quick steps to shader analysis and problem detection with Mali Offline Compiler
    • Q&A: your questions for presenters
    This webinar is part of a series and please don't forget to register for the other sessions by clicking on the attachments section and following the links provided.
  • Optimizing Android Graphics Episode 2: Best Practice for Mobile Game Development Recorded: May 19 2021 75 mins
    Joe Rozek, Developer Relations Engineer, Arm
    Join us live for all three sessions to put your questions to the experts.

    This training series for game developers covers the latest techniques for optimizing mobile game performance. It will give you an in-depth training from fundamentals of underlying GPU hardware to API best practices. Hear first-hand advice from optimization experts and architects of the most prevalent graphics processors across Android devices today. By attending all three webinars you will learn tips for creating more efficient graphics, avoiding game lag and excessive battery usage, and how to discover graphics issues quickly using the free tools available in Arm Mobile Studio. The later session will introduce the modern workflows and tools that help you put in practise the principles that you learnt in the earlier sessions.
    At the end of each webinar, our experts will be available for an extended Q&A session where you can ask questions related to your own specific problems.

    This webinar covers the latest best practice recommendations for graphics API development, which when followed correctly ensure the best performance from commonly deployed Mali GPUs. The session will demystify the high-level design principles for how best to drive the hardware. You will also learn expert recommendations for the major types of artistic content, such a textures, buffers, and shaders.

    Content Covered:
    • Mali GPU best practises
    • Principles for frame construction
    • API best practises
    • The right approach for developing content
    • Latest recommendations for shaders
    • Q&A: Your questions for presenters

    This webinar is part of a series and please don't forget to register for the other sessions by clicking on the attachments section and following the links provided.
  • Optimizing Android Graphics - Episode 1: Mobile Graphics Processing Fundamentals Recorded: May 12 2021 72 mins
    Joe Rozek, Developer Relations Engineer, Arm
    Join us live for all three sessions to put your questions to the experts.

    This training series for game developers covers the latest techniques for optimizing mobile game performance. It will give you an in-depth training from fundamentals of underlying GPU hardware to API best practices. Hear first-hand advice from optimization experts and architects of the most prevalent graphics processors across Android devices today. By attending all three webinars you will learn tips for creating more efficient graphics, avoiding game lag and excessive battery usage, and how to discover graphics issues quickly using the free tools available in Arm Mobile Studio. The later session will introduce the modern workflows and tools that help you put in practise the principles that you learnt in the earlier sessions.
    At the end of each webinar, our experts will be available for an extended Q&A session where you can ask questions related to your own specific problems.

    This introductory webinar explores the mobile GPU (Graphics Processing Unit); how it behaves and where you can find them in modern systems. We provide an overview of the rendering pipeline used by OpenGL ES and Vulkan graphics APIs, explaining each pipeline stage. Finally, we will demonstrate how the rendering pipeline translates into GPU hardware, and how shader cores are designed to execute shader programs efficiently.

    Content Covered:
    • Introduction to mass market Android mobile systems
    • Understanding the rendering pipeline
    • GPU architecture overview
    • Get the basics on hardware shader cores
    • Q&A: your questions for presenters

    This webinar is part of a series and please don't forget to register for the other sessions by clicking on the attachments section and following the links provided.
  • How Machine Learning is Enabling New Smart Camera Applications Recorded: Apr 28 2021 34 mins
    Dylan Zika, Product Manager, Arm
    Smart Cameras is one of the fastest growing areas of Machine Learning. Join Arm for a live webinar to hear how smart cameras are being redefined by AI and ML workloads.
    Today many camera systems still rely on human visual inspection to detect, recognize and track objects of interest. While this is cost effective for single cameras systems, enabling Automatic inspections will lower cost in large scale environments. AI algorithms complement a single person’s ability, focusing human inspection on the most critical, camera video streams.
    Key ML algorithms enabling automatic inspection are object detection, and image segmentation to locate, track, and create boundaries.
    Future systems are increasing security and intelligence with emerging algorithms like human pose estimation or enabling voice interfaces on the camera.
    Dylan Zilka from Arm’s Machine Learning Group will show you how to analyze ML models and give guidance on understanding ML system design and requirements for your products.
    Dylan holds a Masters (MSEE) in Computer Architecture from The University of Texas at Austin, with a technical emphasis on software use-cases and hardware design. His technical Master's was supported with key MBA foundations in Entrepreneurship, Marketing, and Technology Strategy. Dylan’s passion for technologies is broad, with talents spanning hardware, software, and business strategy.

    This webinar will cover:
    • Showcase multiple ML smart camera use cases
    • Learn how to analyze ML models
    • Guidance on how to better understand ML system design/ML requirements for their products
  • Computer Vision at the Edge Recorded: Apr 7 2021 56 mins
    Mary Bennion, Arm, Kevin Gordon, VP of AI Technologies, NexOptic and Colin D'Amore, Machine Learning Engineer NexOptic
    Cameras are everywhere: they're in our cars, homes, places of work, and increasingly in our machines & robotics. AI has brought about rapid change and improvements in deep learning algorithms, so it is less about “what” is seen, but “how” it is seen.

    Canadian company NexOptic create differentiated, scalable, and highly optimized AI software solutions for imaging. Join Kevin Gordon, VP of AI Technologies, Colin D’Amore, Machine Learning Engineer, and Mary Bennion, Manager of AI Ecosystem at Arm, for an eye-opening webinar on Computer Vision at the Edge.

    1. Learn how AI/ML is transforming images and video pixel-by-pixel
    2. Every aspect of any image is uniquely perfected and tuned for precise ultra high-quality imaging
    3. How to leverage raw data on the edge, with a demo from NexOptic

    NexOptic Homepage:
    nexoptic.com

    Sign Up Link:
    https://forms.gle/dyMNjPioU5M4DTKF9
  • Model-Based Development Accelerated: Optimized Code Generation for Arm Devices Recorded: Mar 15 2021 43 mins
    Ronan Synnott, Arm, Abhik Dey, Etas & Rob Barnes, Elsense
    The number of control software for safety-relevant applications running on Arm-based CPUs is rapidly growing. Today a vehicle has over 100 million lines of software code that controls its various systems from engine management to atmospheric control, which is 100 times more than in the Space Shuttle and over four times as much as in a Boeing 787.

    Model-based development has been widely adopted to help manage project complexity and simplify traceability, maintainability and conformance to industry and safety standards. Arm has partnered with industry experts ETAS to streamline toolchain integration and enable highly optimized code generation for Arm-based microcontrollers using ETAS ASCET-DEVELOPER models built with the safety-qualified Arm Compiler as supplied with Arm Development Studio.

    Learn more:

    - How ETAS have collaborated with Arm to provide an easy flow for the development of safety-relevant control applications conforming to the established industry standards.
    - How to achieve optimum runtime performance whilst simplifying toolchain integration
    - How to easily evaluate the flow with an example setup created by Elsense a UK based consulting engineering company

    Speakers:

    Abhik Dey, Product Manager ASCET-DEVELOPER at ETAS, has over 13 years of experience spanning various roles in automotive software domain. He also brings knowledge of the accompanying processes, methodology, standards and tools.

    Rob Barnes, Director and Technical Specialist at Elsense Limited, is responsible for automotive transmission and motor control software. He has over 20 years of experience in series production software for Dual Clutch and Automated Manual Transmissions for drivelines such as the Bugatti Veyron.

    Ronan Synnott, Solutions Architect at Arm, educates developers on all aspects Arm’s development tool offerings, ensuring that they select the right portfolio of products for their specific needs, and enabling the most productive use of the tools together.
  • CMSIS: Practical Use in Embedded Software Development Recorded: Feb 18 2021 64 mins
    Vladimir Marchenko, Product Manager MDK Core and Middleware, Arm & Bruno Castelucci, Principle FAE, Arm
    Arm creates standards, tools and design flows that help developers create reliable and secure products faster. Cortex Microcontroller Software Interface Standard (CMSIS) helps to:

    - Improve the software portability and re-usability
    - Move quickly through product development from evaluation platform to bespoke production hardware
    - Easily manage software components over the product lifecycle
    - Optimize the resource use to minimize system cost
  • Model-Based Development Accelerated: Optimized Code Generation for Arm Devices Recorded: Jan 26 2021 44 mins
    Ronan Synnott, Arm, Abhik Dey, Etas & Rob Barnes, Elsense
    The number of control software for safety-relevant applications running on Arm-based CPUs is rapidly growing. Today a vehicle has over 100 million lines of software code that controls its various systems from engine management to atmospheric control, which is 100 times more than in the Space Shuttle and over four times as much as in a Boeing 787.

    Model-based development has been widely adopted to help manage project complexity and simplify traceability, maintainability and conformance to industry and safety standards. Arm has partnered with industry experts ETAS to streamline toolchain integration and enable highly optimized code generation for Arm-based microcontrollers using ETAS ASCET-DEVELOPER models built with the safety-qualified Arm Compiler as supplied with Arm Development Studio.

    Learn more:

    - How ETAS have collaborated with Arm to provide an easy flow for the development of safety-relevant control applications conforming to the established industry standards.
    - How to achieve optimum runtime performance whilst simplifying toolchain integration
    - How to easily evaluate the flow with an example setup created by Elsense a UK based consulting engineering company

    Speakers:

    Abhik Dey, Product Manager ASCET-DEVELOPER at ETAS, has over 13 years of experience spanning various roles in automotive software domain. He also brings knowledge of the accompanying processes, methodology, standards and tools.

    Rob Barnes, Director and Technical Specialist at Elsense Limited, is responsible for automotive transmission and motor control software. He has over 20 years of experience in series production software for Dual Clutch and Automated Manual Transmissions for drivelines such as the Bugatti Veyron.

    Ronan Synnott, Solutions Architect at Arm, educates developers on all aspects Arm’s development tool offerings, ensuring that they select the right portfolio of products for their specific needs, and enabling the most productive use of the tools together.
  • Arm Cortex-A78AE: On the Road to an Autonomous Future Recorded: Dec 16 2020 31 mins
    Srikanth Rengarajan, Senior Product Manager, Arm
    The Cortex-A78AE is Arm's highest performance CPU designed for devices undertaking complex and demanding safety-critical tasks. Join Arm senior product manager Srikanth Rengarajan as he discusses the key innovations including:

    - A 30% performance uplift compared to the Cortex-A76AE.
    - Features to achieve the relevant automotive and industrial functional safety standards, ISO 26262 and IEC 61508 for applications up to ASIL D / SIL 3. 
    - New enhanced Split-Lock functionality (Hybrid Mode) to offer maximum flexibility. Hybrid Mode is designed to specifically enable applications that target lower levels of ASIL requirements without compromising performance and allow the deployment of the same SoC compute architecture into different domain controllers.
Diverse webinars on all elements of the Arm ecosystem
From providing the IP for the chip to delivering the cloud services that allow organizations to securely manage the deployment of products throughout their lifecycle, Arm delivers a complete Internet of Things (IoT) solution for our partners and customers. It’s rooted in our deep understanding of the future of compute and security.

Embed in website or blog

Successfully added emails: 0
Remove all
  • Title: Bringing PyTorch Models to Cortex-M Processors
  • Live at: Jun 28 2021 12:20 pm
  • Presented by: Rohit Sharma, AITS
  • From:
Your email has been sent.
or close