Only two of the top five public cloud providers, AWS, Azure, Google Cloud Platform, IBM Cloud, and Alibaba Cloud, have a fully developed edge computing strategy. There are still a number of players who haven’t figured out their edge computing strategy.
A service called AWS Greengrass, first launched at Amazon’s 2016 re:Invent event, is used to provide access to Amazon’s edge platform. AWS just added the ability to execute inferencing of machine learning models. It has begun incorporating AWS Greengrass in products like as AWS DeepLens, a smart camera that can run neural nets at the edge.
Almost a full year after AWS Greengrass’ launch, Microsoft released their own IoT platform, Azure IoT Edge. However, the wait has been well rewarding. To begin with, the last twelve months have seen a shift in market dynamics, providing an opening for the group to better accommodate client needs. Secondly, Microsoft received a chance to improvise its platform to make it better than the sole alternative option – AWS Greengrass.
The very first edition of Azure IoT Edge seems to be comprehensive. Here are 5 reasons why I think Microsoft got its edge computing approach right:
1. Making the code for the system freely available
Azure IoT Edge is accessible as an open source project on Github.
Edge computing use cases are continually growing where clients are expected to use the platform in new ways. To promote flexibility and openness, Microsoft has handed up the source code of its edge computing platform. Through this, users will be able to personalise their deployments based on Azure IoT Edge. Modifying the Azure IoT Edge source code enables the addition of historical protocols, integration with existing asset management solutions, and interoperability with proprietary communication protocols and data formats.
Microsoft’s decision to release its edge software under an open source licence was inspired. It simply improves the trust and confidence of customers.
2. Containers at the core
Microsoft’s Azure IoT Edge is powered by Moby, the same open-source container runtime that drives Docker. Due to this architectural choice, developers may now package and deploy modules built on top of regular Docker containers on Azure IoT Edge.
Docker containers can be chained together like UNIX Pipes, with the output of one module serving as the input for another. Containerized components of Azure services including Stream Analytics, Functions, and SQL Server are being made available by Microsoft for use at the edge.
It is not necessary to deploy the full programme in order to manage and update individual modules. The container images can be kept in a cloud-based or on-premises registry. Customers can set up CI/CD pipelines to have the most recent module versions (container images) pushed out to various edge nodes automatically. To further orchestrate the distributed edge deployments, Microsoft is also looking into integrating Kubernetes with Azure IoT.
With containers, it’s much easier to manage sophisticated, dispersed edge deployments, as both Azure services and custom logic can be packaged and shipped in a single unit.
3. Ecosystem engagement
The robust ecosystem of original equipment manufacturers (OEMs) and original design manufacturers (ODMs) supporting Azure IoT is now being rolled out to Azure IoT Edge. The IoT Edge certification programme has the capability-based certification paradigm. The certification programme assigns a certain tier to each vendor that participates in order to clearly define their abilities. A vendor whose focus is on the core runtime would receive the first tier, while a vendor with a strong security offering could reach the fourth tier. The diverse network of partners supplying edge computing solutions is made available to clients through this capability-based marketplace.
Microsoft has also added support for Azure IoT Edge to the Visual Studio Team System and Visual Studio Code IDEs. Programmers can jump right into constructing the modules by using one of the many available pre-defined templates. Customers may manage the entire module lifecycle using CI/CD pipelines built in to VSTS.
When thinking about the Azure IoT platform, Azure IoT Edge makes perfect sense. To safely provision thousands of devices, it uses services like Device Provisioning Service. The embedded Security Manager abstracts the secure silicon hardware to provide a safe environment for the entire IoT Edge device and its constituent parts. When necessary, ODMs can strengthen the system with HSMs (HSM).
Reasoning Machines on the Fringe of Artificial Intelligence
The use of machine learning models at the edge is now simplified by Microsoft. Each inferencing model may be distributed as a standalone component. On Azure, programmers can use Data Science Virtual Machines or Azure Machine Learning Studio to train their models. Models exported from Azure AutoML services, such custom vision, can be executed on Azure IoT Edge as well. Due to the fact that each model is merely a container/module, new models can be rapidly distributed to the forefront.
Microsoft’s commitment to ONNX paves the way for ML models developed in various frameworks to be converted to a common format and then used for inference.
Microsoft’s plan to create an Intelligent Cloud and Intelligent Edge depends heavily on Azure IoT Edge. Azure IoT Edge is one of the most complete edge computing platforms on the market with to design decisions including containerized modules, strong connection with HSM, and plugins for Visual Studio.
I’d appreciate it if you’d follow me on Twitter and/or LinkedIn. To learn more, visit my online portfolio.
We thank Janakiram MSV
M.S. Janakiram is an analyst, advisor, and architect at his own firm, Janakiram & Associates. In addition to starting Get Cloud Ready, he also served as its chief technology officer.
Policies on Reprints and Permissions from the Editors
It’s possible that the Orioles are a better investment than the Nationals, given Juan Soto’s contract rejection.
Staff Writer, Forbes Mike Ozanian, on the convergence of business and athletics
Today is July 18th, 2022 at 9:08am EDT.
Nationals of Washington vs. Reds of Cincinnati
Juan Soto is confident that he can increase the value of his company by more than $400 million in 15 years. Possibly he can do it. Photo by DYLAN BUELL/GETTY IMAGES
Juan Soto allegedly rejected a 15-year, $440 million contract offer from the Washington Nationals over the weekend, potentially shifting the dynamics of which baseball franchise in the area—the Nationals or the Baltimore Orioles—is the better buy.
The Lerner family, owners of the Nationals, are reportedly seeking $3 billion. The Orioles, who play an hour north of Nationals Park, may soon be up for sale as well, due to internal strife among the family that has long controlled the organisation. Most recently, Louis Angelos filed a lawsuit against his sibling John and their mother Georgia.
Forbes estimated the Nationals to be worth $2 billion in March, while the Orioles were worth $1.38 billion. The Nationals have been a far stronger team than the Orioles during the past decade, culminating in their 2019 World Series championship. That’s why the Nationals have been bringing in roughly a third more fans than the Orioles. The Nationals made $322 million that year, whereas the Orioles only made $251 million.
It’s a different story on the field, where the roles of the two teams have been reversed this year. While the Nationals are 19 games out of the National League playoff picture, the Orioles are playing at a.500 pace and are only 3.5 games shy of a wild-card place in the American League. The Nationals spend $135 million more than the Orioles on payroll, whereas the Orioles only spend $45 million. The Nationals have a history of delaying contract signings, which is even worse. In comparison to the Nationals’ $29.5 million, the Orioles’ delayed or retained (buyouts or releases) salary this season are only $15.8 million. And in 2023, the Nationals will be responsible for $23 million in deferred contracts, while the Orioles will owe only $15 million.
As a result, a new owner for the Orioles would have far greater financial leeway than the Nationals. If Soto were to leave for another team, the Orioles’ current condition on the field would likely improve.