Arm TechCon will take place on October 8-10, 2019 at San Jose Convention Center to showcase new solutions from Arm and third-parties, and the company has now published the agenda/schedule for the event.
There are many sessions and even if you’re not going to happen it’s always useful to checkout what will be discussed to learn more about what’s going on currently and what will be the focus in the near future for Arm development. Several sessions normally occur at the same time, so as usual I’ll make my own virtual schedule with the ones I find most relevant.
- 09:00 – 09:50 – Open Source ML is rapidly advancing. How can you benefit? by Markus Levy, Director of AI and Machine Learning Technologies, NXP
Over the last two years and still continuing, machine learning applications have benefited tremendously from the growing number of open source frameworks, tools, and libraries to support edge inferencing. These include CMSIS-NN, ARM NN, TensorFlow Lite, OpenCV -a few of the more popular options. This session begins with an analysis of the criteria for selecting the most appropriate technology (e.g. performance, maturity, frameworks supported). The biggest benefit of open source is that it ‘rises the tide’ for all and can rapidly improve because of the on-going community support, and hence, the next topic for this session is an exploration of the current state of the art for open source and how to enable it. But things get even more interesting when users can add compute-specific optimizations on top of the open-source technology to increase performance. This brings us to the session’s conclusion which focuses on these optimizations and how to deploy them on Arm-based MCUs and MPUs.
- 11:30 – 12:20 – Practical steps for developing secure IoT endpoints by Christopher Seidl, Senior Marketing Manager, Arm
This sessions demonstrates a practical approach for developing secure IoT endpoints from a software developer’s perspective. Following Platform Security Architecture (PSA) principles, this talk walks attendees through typical design phases and discusses how to apply standardized software solutions such as Trusted Firmware-M, CMSIS components and Keil MDK to efficiently tackle common challenges.
- 13:30 – 15:30 – Develop, connect, and manage IoT devices with Pelion Device Management and Mbed Studio by Arkadiusz Zaluski , Senior Tools Development Engineer, Arm
At this workshop, you will find out just how easy IoT development can be by using Mbed Studio, the brand new IDE from Arm, to develop an Mbed OS application and then connect a device to the Pelion Device Management services. Delegates will have the opportunity to get hands-on with the intuitive and straightforward development workflows and tools offered by Mbed OS and Pelion services while learning useful hints from the creators themselves. The session will also explore the deep layers of security baked into Arm’s IoT products, and provide the opportunity to join our early access program so that you can help to guide the future of IoT tooling.
- 15:30 – 16:20 – Real-time dataflow: bridging DSP and ML for embedded systems by Glenn Kasten, Software Engineer, Google
A dataflow software architecture models computation as a directed graph, where the nodes are pure functions, and the edges between nodes are data. In addition to recent uses in deep learning, big data, and reactive programming, dataflow has long been an ideal fit for Digital Signal Processing (DSP). In a sense, artificial neural networks can be thought of as DSP with large adaptive filters and non-linearities. Despite the success of dataflow in deep learning and DSP, there has not yet been to our knowledge a lightweight dataflow library that fulfills these requirements: small (under 50 Kbytes code), portable with few dependencies, open source, and most important: predictable performance suitable for embedded systems with real-time processing on the order of one millisecond per graph evaluation. We describe a real-time dataflow architecture and initial C++ implementation for Arm processors that meet these requirements, then explore the benefits of a unified view of ML and DSP.
- 16:30 – 17:20 – Arm-based Linux IoT: from prototype to production by Drew Moseley, Technical Solutions Engineer, Mender.io
We will discuss some of the considerations device manufacturers should consider when designing Linux-based connected devices. These devices are increasingly common in the Internet of Things. We will discuss hardware, software, security, and how to bring it all together. We will present a demo solution using a Raspberry Pi device and provide a build environment and instructions for attendees to use on their own hardware.
- 09:00 – 09:50 – A demo: The Cloud, Edge, and Internet of Things… All running on Arm! by David Tischler, Founder, miniNodes.com and Carl Perry, Ecosystem Engineer, Packet
In the weeks leading up to Arm TechCon 2018, and then at the event, Drew Henry outlined his vision of a trillion connected devices. Those trillion connected devices will need a significant rethinking of how infrastructure is built and delivered, and the Arm Neoverse initiative was created to address those challenges. The past 6 months have seen much written on the topic of bringing workloads back from the Cloud, and moving them to the Edge, closer to the end users or to IoT endpoints, and improving the service delivery experience. While there have been many articles, slides, headlines, and conversations about this, no one has yet to demonstrate a full end-to-end working implementation. miniNodes is building a complete demonstration of connected Cloud Servers, Edge Servers, and IoT Devices, running entirely on Arm. Data will be captured by IoT endpoints running Arm Mbed, provisioned via Arm Pelion, feeding data to Edge servers, that will in turn connect to an AWS A1 Arm Server.
- 11:30 – 12:20 – Arm NN deployments in edge devices by Andrea Gallo, VP of Membership Development, Linaro
Arm and Linaro launched the AI initiative one year ago to collaborate on an open source inference engine common to all Arm edge devices and support SoC specific NN acceleration via a plug-in back end framework. The mlplatform.org platform hosts the upstream open source work for both Arm NN and the Arm Compute Libraries. The team, made up of engineers from Arm, Linaro, Qualcomm, TI, and other members, is deploying Arm NN in edge devices through the integration in upstream projects like Tensorflow Lite and AWS TVM. Andrea Gallo, Linaro VP of Membership Development, will provide an overview of the ongoing activities to add support for multiple SoCs in Arm NN, set up CI and testing infrastructure, integrate in runtime frameworks and graph compilation technologies.
- 13:30 – 15:15 – Build a low-powered Arm voice assistant with Google TensorFlow Lite by Alessandro Grande, Ecosystem Manager, Automotive & IoT, Arm, Peter Warden, Staff Research Engineer, Google, and Wei Xiao, Principal Developer Ecosystem Evangelist, Arm
The world we live in today is full of computers we cannot see. Last year alone, Arm and its partners shipped 23 billion processors. Most of these processors did not go into laptops or mobile phones, but rather in objects that we don’t consider as traditional computers such as toothbrushes, wearables, speakers, or factory equipment. Today, it’s hard to find an object that does not contain a processor. What if all these smart devices allowed everyday objects to become aware of the environment they are in? Advances in processing power and machine learning algorithms are enabling more of the computing to happen on edge devices. The collaboration between Arm and Google is enabling the tiniest of these devices, microcontrollers with only kilobytes of memory, to become smarter. This workshop will be a hands-on session in which you will become familiar with the end to end flow necessary to develop a keyword spotting application, optimized to run on an ultra-low-power Arm Cortex-M4 processor. TensorFlow Lite Micro together with a library optimized for Cortex-M microcontroller
- 14:30 – 15:20 – Managing security of Connected IoT Devices – Do nothing or do it RiGHT! by Sameer Dixit, Vice President, Security Consulting, Spirent Communications
Internet of Things (IoT) deployments have been growing at an astonishing pace – whether for data monitoring, facility management, manufacturing processes or supply chain. However, there continues to be a gap in understanding the potential risks, and hidden threats that exist. This presentation will describe the current state of IoT security, provide insights into the ever-evolving world of hidden threats, identify latest IoT cyber-security standards and its contribution in enhancing overall platform security. Attendees will also receive guidelines on securing various components of an IoT deployment.
- 15:30 – 16:20 – IIoT: Security in smart manufacturing by Brian Clinton, Senior Director Engineering Services, Arm and Bryan Zhang, Staff Security Consultant, Arm
Industrial IoT is going through digital transformation. A smart factory is highly digitized and connected, such data-driven autonomous systems and machine learning capabilities of the edge bring significant operational benefits. However, it also brings security challenges due to much more attack surfaces becoming exposed as a result the manufacturing systems being connected and managed from the cloud.
- 09:00 – 09:50 – Designing with Arm Cortex-M1 IP to turn an FPGA into a low cost μSoC FPGA by David Grugett, Sr. FAE Manager, Gowin Semiconductor
Arm’s DesignStart FPGA program is a great way for embedded designers to create low-cost µSoC FPGA solutions. The no-fee license and royalty-free Cortex-M1 Controller gives developers an eco-system with the broadest of software and tools to design and innovate. Coupled with FPGAs, one can develop complex applications that include off-load acceleration, extended compute functionality, and always on/low power solutions. By choosing Arm Cortex M1 for FPGA, designers can accelerate success in designing reliable, low-power, and secure solutions. This hands-on demonstration of Arm’s DesignStart FPGA workflow on a GOWIN FPGA platform will illustrate key principles that engage the audience in wanting to use µSoC FPGAs for video interfaces, smart connected devices, portable consumer, IoT, AI and Edge computing solutions.
- 11:30 – 12:20 – Optimizing deep learning for applications on microcontrollers by Laurent Folliot, Director, Machine Learning Software, ST Microelectronics, Matthieu Durnerin, Director, ML Software, ST Microelectronics, and Rob Elliott, Director Applied Machine Learning, Arm
Complex and various Deep Learning Frameworks have appeared, for example TFLite for Microcontrollers, and the need for tools capable of handling such an ecosystem for various applications has emerged. In this session, we will present Neural Network Quantization, one of the latest and most promising evolution for constrained-resource devices. We will illustrate quantization results and performance on various applications with Cube.AI tool taking also advantage of CMSIS-NN, and latest Deep Learning frameworks evolutions.
- 13:30 – 14:20 – When it comes to connect IoT devices, how small is small? by Derek Atkins, Chief Technology Officer, SecureRF Corporation
IoT developers are familiar with TLS-based solutions that enable devices to communicate securely over a network. But because TLS has its roots in enterprise computing where processing resources abound, it was not designed to run on constrained platforms. There are IoT frameworks that leverage TLS, but those exclude lower resource processors because TLS will not fit. We present an alternative to TLS’s ECC and RSA based on Group Theoretic Cryptography (GTC) designed specifically for low resource processors. GTC enables quantum-resistant drop-in replacements for ECDH and ECDSA that allow a low resource processor to securely communicate with gateway devices over an open channel. We present a real-world application that shows how a Cortex-M0-based MCU provides anticounterfeiting plus data protection for a consumable item within an appliance. The Cortex-M0 transmits analytics to its host appliance then on to a cloud server without running on top of an OS or employing TLS.
- 14:30 – 15:20 – The security mindset for cloud connected medical devices by Kaushal Vora, Marketing Director, Healthcare & Emerging Technologies, Renesas Electronics America Inc. and Onkar Raut, Staff Applications Engineer, Healthcare & Emerging Technologies, Renesas Electronics America Inc.
High-performance MCUs with large memory, supporting IP, and cloud connectivity are transforming the medical market, enabling the deployment of complex software applications into critical infrastructures. What used to be simple point-solutions to address a single problem now collect and transmit data for better insight into the problem, and provide recommendations to users & developers to improve the solution itself. Requirements for modern-day medical devices, like connectivity, are often added onto the device’s core functionality. However, establishing connectivity without considering the threat landscape can be disastrous, especially if a security breach compromises safety. This session will analyze common operation risks, the impact of regulatory guidelines, and key hardware and software components needed to maintain safe and secure operation. The session will also show how a platform approach can minimize operation risks while helping developers speed their applications to market.
- 15:30 – 16:20 – Turn soft spots into guardian: Blockchain technology for edge devices by Mei-Ling Chiang, Sr Staff Product Marketing Manager, Marvell and Trent Poltronetti, VP Sales and Marketing, SmartAxiom Inc.
With billions of devices connecting to the cloud and artificial intelligence pushing to the edge, centralized security starts to fail as keys and certificates become unmanageable and latencies rise. What’s worse, remote servers and their network connections typically lack redundancy and are susceptible to man-in-the-middle attacks. Instead, decentralized blockchain technology can enable fast, local decisions while improving system security. It can also bring more intelligence to the edge with smart contracts. While blockchain was initially used to build trust in digital currencies, when scaled appropriately its security architecture is ideal for edge devices. Learn why decentralized security is so critical in future edge devices and how Marvell embeds blockchain security powered by SmartAxiom to secure billions of edge devices, improve efficiency and reduce cost of ownership.
Many of the sessions above deals with either artificial intelligence / machine learning, or IoT security, but there are also a fair amount of talks about containers (e.g. Docker), chip design, automotive applications, and more in the full agenda.
If you’ll like to attend Arm Techcon 2019, you’ll need to register, and purchase a pass/ticket, unless you only plan to attend the exhibition hall. Here are the options:
- $999 All-Access Pass ($550 off onsite price) – Access to visionary keynotes, workshops and sessions across seven conference tracks, and the Expo Hall.
- $599 Alumni All-Access Pass ($950 off onsite price) – Available to anyone who has attended Arm TechCon with an All-Access Pass or Expo Pass in 2014 or thereafter.
- $299 Student Pass – All-Access Pass for teachers and students that register with a .edu email and bring valid teacher or student ID upon check-in.
- $499 Government Pass – All-Access Pass for government employees that register with a .gov email and bring valid government ID upon check-in.
- Free Expo Hall Pass – Keynotes, leading technology suppliers, industry education and networking. (Note: October 9-10 only)
Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.