March Newsletter

Welcome to this edition of the newsletter. Because of the length of this edition, I have embedded two podcasts, one for each of the events covered.

Events

There were a few technology events in March and I will provide some information on two of those.

Nvidia GTC 2023

The Nvidia GPU Technology Conference took place over three days, starting from March 20th. The focus of the conference was on sustainability, Generative AI, and industrial digitization.

You can watch a video podcast of this section here.

You can listen to an audio podcast of this section here.

In recent years, sustainability has become a top priority for many industries. With the increasing demand for computational power, it's no surprise that companies are seeking ways to reduce their carbon footprint while still meeting their computing needs. By shifting computationally intensive tasks to GPUs, companies can realize significant energy savings and reduce their environmental impact. Here are some additional points to consider:

• GPUs are designed to handle parallel processing, which makes them ideal for computationally intensive tasks like machine learning and data analysis.
• Compared to traditional CPUs, GPUs can process data much faster while consuming less power, which can result in lower operating costs and a smaller carbon footprint.
• In addition to energy savings, using GPUs can also improve overall system performance and reduce processing times, which can lead to increased productivity and faster time-to-market for products and services.
• While there may be some initial costs associated with upgrading to GPU-based systems, the long-term benefits in terms of energy efficiency and performance can make it a worthwhile investment for companies looking to reduce their environmental impact.

Overall, by accelerating industrial workloads with GPUs, companies can achieve both their computing needs and sustainability goals.

The updates to Nvidia's Omniverse platform are expected to have a significant impact on the future of metaverse applications. Here are some additional details to provide further context on this exciting development:

• With the updated platform, users can create and operate metaverse applications that are more realistic and dynamic than ever before.
• The use cases for these applications are broad and varied, with companies in industries ranging from robotics to automotive utilizing the technology.
• One key advantage of Omniverse is its ability to create digital twins that can be used to simulate real-world scenarios in a virtual environment. This can lead to more efficient testing and development processes across a range of industries.
• The platform also offers enhanced collaboration features, allowing teams to work together on complex projects seamlessly.
• With the ongoing growth of the metaverse, it's clear that Nvidia's Omniverse platform is well-positioned to be a key player in this exciting new space.

Nvidia has unveiled Nvidia Omniverse Cloud, which will be hosted on Azure and closely integrated with Azure IoT and the Office productivity suite.

They also announced updates to a large number of their libraries and SDKs.

If you haven't already, I would recommend taking a look at Nvidia LaunchPad. It's a platform that lets you take free, hands-on labs running on Nvidia servers. You will need to apply to take a lab, but once approved you will have access for a period of two weeks.

For the data center, Nvidia announced new chips: a new Grace CPU, a Grace SuperChip that has two onboard Grace CPUs, and a Grace Hopper SuperChip that has a Grace CPU and a Hopper GPU onboard.

The DGX family of supercomputers are Nvidia's solutions for Deep Learning workloads. However, these have always been expensive. Nvidia has announced the DGX Cloud as a way to make access to Deep Learning infrastructure cheaper and easier. In their words, it's "AI Training-as-a-Service", with a promise of predictable monthly pricing. It's coming first to the Oracle Cloud Infrastructure but will come to Microsoft Azure and GCP as well.

The DGX Cloud will prove useful to companies with all of the resources to train and deploy models from scratch. Nvidia AI Foundations is a new service that will provide pre-trained models for companies who want to fine-tune (build, refine, operate) models for Generative AI. There are three classes of models available: language models, vision models, and biology models.

Finally, they announced the Nvidia Inference Platform. This is made up of four new servers: the L4 for AI video workloads and available on GCP, the L40 for Omniverse and Generative AI inferencing, the H100NVL for Large Language Model (LLM) Inferencing, and the Grace Hopper for Vector Databases, Generative AI & Recommenders.

Arduino Day 2023

Arduino Day is an annual celebration of the Maker Community, and Arduino's opportunity to announce new products and services. It took place on March 25th. There was a large community showcase but I will skip over those and focus on Arduino's announcements.

You can watch a video podcast of this section here.

You can listen to an audio podcast of this section here.

Arduino unveiled its latest lab highlight: Arduino lab for MicroPython – a new IDE that's simpler than the previously provided OpenMV. This new IDE supports all of Arduino's boards that are compatible with MicroPython - including Nano RP2040 Connect, Nano 33 BLE Sense, Portenta H7, Nicla Sense ME, Nicla Vision, and Nicla Voice.

Arduino Cloud is a platform for developing, deploying, and monitoring devices. It's made up of two solutions: Web Editor for writing and uploading code, and IoT Cloud for managing devices holding data, and building dashboards. Arduino has announced support for new hardware, including their own Mega, Opta, and Portenta H7, as well as ESP32-C3, ESP32-S3, and ESP32-S3 from Espressif Systems. Arduino Cloud now has support for Over-the-Air (OTA) updates for Espressif boards.

You can now utilize your phone as a device on Arduino Cloud, providing access to sensors on your phone. This integration between your phone and Arduino Cloud opens up a whole new world of possibilities for your projects. Here are some ways you can use your phone's sensors in your Arduino projects:

• Use the accelerometer to detect changes in motion and orientation
• Utilize the GPS to track your asset's location and movements
• Use the compass to determine which direction your project is facing
• Use the barometer to measure atmospheric pressure and altitude

With these additional sensors, your Arduino projects can become even more sophisticated and responsive to their environment. Plus, the ability to use your phone as a device on Arduino Cloud makes it easier to rapidly prototype without waiting to get new hardware and sensors.

With the Arduino Cloud CLI, managing IoT devices has never been easier. Here are some additional benefits and use cases for the CLI:

• The Device Provisioning feature allows you to easily add new devices to your IoT network. You can provision your devices in bulk, making it easy to expand your network as your needs grow.
• Thing Management allows you to view and manage your devices in a centralized location. You can easily monitor device health, view device data, and troubleshoot issues in real time.
• The Dashboard Management feature allows you to create custom dashboards to visualize your device data. You can easily create charts and graphs to display important metrics and insights.
• The Export text definitions feature allows you to export your device configurations as templates, making it easy to replicate your network across multiple devices.
• You can easily push firmware updates to your devices over the air. This saves time and resources, as you can update multiple devices at once without having to physically access each one.

Overall, the Arduino Cloud CLI provides a comprehensive set of tools and features that make managing IoT devices simple and efficient.

The features offered by Arduino Cloud for Business can greatly benefit enterprise clients who are looking for efficient device and fleet management solutions. Some additional context around these features could include:

• Device Management: This feature allows businesses to easily monitor and control their devices remotely, including updating firmware and managing configurations.
• Fleet Management: With the ability to manage a large number of devices, businesses can streamline their operations and reduce the risk of downtime.
• Role-based access control: This feature ensures that users only have access to the features they need, which can help prevent unauthorized access and enhance security.
• Optional add-ons: The Portenta X8 Manager is a great tool for businesses that have deployed the Portenta X8, as it provides additional capabilities and flexibility. Additionally, the Enterprise ML Tool offers businesses the ability to train and deploy custom ML models, which can help improve operations and drive innovation.

Overall, Arduino Cloud for Business offers a comprehensive suite of tools and features for enterprise clients, making it a great choice for businesses looking to optimize their device and fleet management processes.

The Arduino Project Hub has undergone a major overhaul to make it more user-friendly and accessible. Projects are now categorized and trending projects are highlighted. You can even find templates to help you create new projects. The code section is now integrated into the Arduino Cloud, making it easy for others to replicate your code. With the new and improved Arduino Project Hub, you can trust it to manage your project portfolio.

There is a new Project of the Month Competition that awards prizes to as many as three projects every month. You will need to submit your project using an individual account, and not a company account. Also, you can't submit projects using Hackster. Winners will be awarded gift cards that they can redeem on the Arduino Store.

Arduino highlighted the Nicla family of thumb-sized microcontrollers:

• The Nicla Vision is equipped with a camera for computer vision applications, as well as a microphone, distance sensor, and a 6-axis IMU.
• The Nicla Voice is equipped with a microphone for speech recognition applications, as well as a 6-axis IMU.
• The Nicla Sense ME is equipped with motion and environment sensors.

All three Nicla microcontrollers are equipped with processors that are capable of running Machine Learning models. They also have WiFi and Bluetooth Low Energy (BLE) connectivity.

Arduino also highlighted the Arduino Opta microPLC with industrial IoT capabilities. The Opta features the same processor as the Portenta H7, giving it TinyML capabilities as well. There are three variants, with each one having progressive connectivity.

• Opta Lite has ethernet connectivity.
• Opta RS485 adds Modbus connectivity to the base Opta Lite.
• Opta WiFi adds WiFi and BLE to the Opta RS485.

The Opta can be programmed using either the traditional Arduino IDE or the new Arduino PLC IDE.

Finally, Arduino Announced the Uno R4, their newest microcontroller based on the Arduino Uno form factor. More on this further down.

Reference Designs

Accelerators from Blues Wireless are a great resource for developers who want to create solutions using Notecard and Notehub. Here are some additional points to consider:

• One of the key benefits of Accelerators is that they allow developers to quickly prototype and test a wide range of IoT solutions.
• The Accelerators cover a diverse range of use cases, from monitoring power quality and indoor air quality to tracking cargo location and vehicle telemetry.
• By using the Accelerators as a starting point, developers can save time and effort in building and testing their solutions.
• In addition to the Accelerators, Blues Wireless provides comprehensive documentation and support to help developers get the most out of the Notecard and Notehub.

Embedded Systems

Arduino Unveils Portenta C33 with Arm Cortex-M33 Processor for Enhanced IoT Security?

The Portenta family has just expanded with the latest addition – the Portenta C33. Unlike the previous models that featured the dual processor STM32H747 with an M4 and an M7 processor, this new member boasts an Arm Cortex-M33 processor from Renesas specifically designed for IoT and embedded applications that require optimal security. Using Arm's V8 architecture, the Portenta C33 is equipped with Wi-Fi and Bluetooth Low Energy capabilities.

Introducing the Latest Addition to the Arduino Family: Arduino Uno Rev 4?

Arduino has just launched their newest microcontroller development board, the Arduino Uno Rev 4. With the same form factor as its predecessor, this board is designed to ensure continuity and compatibility with shields. Additionally, it operates on 5V power, so previous project peripherals will continue to function. The Uno Rev 4 comes equipped with USB-C and a 32-bit Arm Cortex-M4 processor from Renesas. This release will include two board versions: the Uno R4 WiFi with onboard connectivity, and the Uno R4 Minima. The Uno R4 will be released at the end of May.

If you have written libraries for the Uno, or you are a contributor to an open-source project that utilizes the Arduino Uno, you are qualified for the Early Adopter Program. This program will grant you access to an Uno R4 before the official release. More about this here.

Introducing the Arduino Giga R1: A New Addition to the Mega Family?

Arduino has just revealed that it's adding another member to the Mega line – the Arduino Giga R1. The new board has the same form factor as the Mega and Due models but is equipped with a powerful new dual-core 32-bit brain, the STM32H747XI from ST Microelectronics. Additionally, it features Wi-Fi and Bluetooth capabilities, making it compatible with Arduino Cloud. Excitingly, this board can operate on a voltage range from 6-24V.

Nvidia Unveils Jetson Orin Nano Developer Kit for Enhanced Entry-Level Edge AI Applications

?The Jetson Orin Nano Developer Kit by NVIDIA is a game-changer for creating AI-powered robots, smart drones, and intelligent cameras, all while streamlining the process of getting started with the Jetson Orin Nano series. This innovative kit sets a new standard for entry-level robotics applications.

Introducing Xiao SoM: A New Way to Decrease Development Time and Costs?

Seeed Studios has recently unveiled the Xiao SoM, an innovative solution that helps designers and makers cut development time and costs by easily integrating it into their products. The Xiao family of thumb-sized microcontrollers comes with castellated pins that support pin soldering or surface mounting on a larger carrier and has various architectures.

Tools

STM32 VS Code extension

Sony Spresense SDK 3.0.0

Machine Learning

KerasCV is a collection of modular building blocks that enable computer vision engineers to rapidly construct production-level training and inference pipelines.

KerasNLP is a powerful natural language processing library that provides full support throughout the development cycle. It features workflows that are constructed from modular components that come with state-of-the-art preset weights and architectures, making them easily customizable when more control is required.

Databases

AlloyDB: The Ultimate Managed PostgreSQL-Compatible Database Service

?Need a top-notch database service that can handle even the most demanding enterprise workloads? Look no further than AlloyDB. This fully-managed solution combines the power of Google with PostgreSQL, resulting in unparalleled performance, scalability, and availability.

Upcoming Events

Getting Started with Blues Wireless and Notecard - April 13th

Workshop on the Scientific Use of ML on Low-Powered Devices - April 17th - 21st

Thank you for reading, until next time.