What is edge computing?
Edge computing refers to a solution where data processing, analysis and in some cases, actions, occur close to the place where the data originated. Edge computing often relies on a sporadic connection to cloud computing systems, although some setups similarly connect to nearby devices — in which case the systems might be referred to as part of the Internet of Things (IoT). Edge computing solutions operate in circumstances where current cloud computing systems won’t suffice, due to one or more of the following concerns:
- large quantities of data
- network connectivity constraints
- network latency limitations
- data privacy priorities
Wherever you encounter one or more of the above four constraints, you’ll also find an example of an edge computing solution. Machines, such as autonomous cars or industrial robots, generate huge quantities of data and act with low latency. Some agricultural systems operate in areas that lack high-bandwidth network connections. Within buildings, various environmental, security and transit systems all benefit from the ability to sense, analyze and respond immediately, independent of cloud systems. And in healthcare, many organizations prefer to maintain as much patient data on-site as possible.
Structurally, you might think of edge-cloud computing as a modern incarnation of client-server computing. The two systems share a common concept: An independent, capable computing device (i.e., the legacy client or modern edge device) that can connect to a more powerful system (i.e., the legacy server or modern cloud). Of course, the capabilities, number, scale, and details of client-server and edge-cloud systems differ, but that general concept provides a useful initial framework as you seek to understand the field.
SEE: How to implement edge computing (TechRepublic)
Companies apply the term “edge computing” to refer to a diverse set of networking technologies, devices (processing and storage systems), applications and data systems. Edge computing solutions from established companies tend to reflect each vendor’s established expertise: Chip companies promote processing, networking companies discuss connectivity, software vendors point to machine learning and development platforms, and hardware makers emphasize devices. But even within these categories, offerings vary widely: Edge computing devices, for example, run the range from small and IoT devices to relatively robust cloud-system-in-a-box data center devices.
The companies covered below represent only a small sampling of the hundreds of companies active in the edge computing space. I placed each of the selected companies into categories — cloud, processors, equipment & integration and software & platforms — to provide some indication of each company’s broad area of strength, although some companies compete in several of these categories. The companies below are ones I encourage you to watch as edge computing solutions evolve.
Cloud companies extending to edge computing
The following three dominant cloud computing vendors also each seek to extend their reach to places not presently served by their existing data centers.
Amazon Web Services
As a dominant competitor in the world of cloud computing, AWS edge solutions extend the company’s cloud expertise into hybrid cloud (e.g., AWS Outposts), edge networking (AWS Wavelength and AWS Local Zones), and industrial and IoT environments. AWS offers edge devices that collect and process data, such as the AWS Snow Family with storage capacities from 8 TB to 100 PB and in sizes that range from roughly comparable to a large book to the scale of a shipping container. Follow @awscloud on Twitter for updates.
Google Distributed Cloud Edge seeks to provide customers the compute power of Google infrastructure and services in any one of four locations: Google’s network edge, an operator’s edge (e.g., communications service provider), customer’s edge, or in a customer’s data center. The solution relies on the company’s multi-cloud management platform, Anthos, and may be deployed either in a configuration that requires a connection to Google Cloud or one that stores and processes data locally with an Edge Appliance. Follow @googlecloud on Twitter for updates.
Azure private multi-access edge compute is how Microsoft refers to a set of services designed to help enterprises manage edge services and applications, networking, compute and radios and devices. Azure private MEC involves many ecosystem partners, in addition to a wide range of core Azure services and Azure Stack Edge managed devices from Microsoft. Separately, Azure IoT offerings include Azure IoT Hub, Azure Digital Twins and Azure IoT Central to help companies manage and automate all sorts of devices and sensors. Follow @azure on Twitter for updates.
SEE: Edge computing: A cheat sheet (TechRepublic)
Processors and edge computing
These three vendors want to provide the processors that operate in edge environments, such as factories, industrial operations, retail environments, smart cities, vehicles and almost everywhere else around the globe.
Intel products and technologies, including processors, of course, are inside all sorts of edge solutions and devices, including computers, sensors, cameras, storage systems, servers, networking equipment, as well as ruggedized and specialized equipment. And while the company encourages software development with open development tools such as the Intel Smart Edge portfolio and Intel Distribution of OpenVINO toolkit for vision applications, its core strength may be how it collaborates with partners to provide Intel Market Ready Solutions and build a broad Intel Network Builders Edge Ecosystem. Follow @inteliot on Twitter for updates.
While NVIDIA has had historic strength in graphic processing units, the NVIDIA EDGE Enterprise platform provides hardware and software systems widely used for artificial intelligence for object or speech recognition and graphical data visualization, among many other applications. The NVIDIA Jetson platform offers modules for autonomous machines and various embedded system uses, with a range of systems that support different processing, environmental durability and power efficiency profiles, with Jetson Software that allows solutions to be developed once and deployed across all Jetson modules. Follow @nvidiaembedded on Twitter for updates.
NXP has a long history of innovation that you might readily recognize under earlier activity as Philips Semiconductors. Their EdgeVerse portfolio includes EdgeVerse Processors, EdgeVerse Microcontrollers, EdgeLock Security solutions and NXP EdgeReady applications (e.g., face recognition and voice recognition solutions). Processors, sensors and edge-to-cloud solutions are especially significant in the company’s offerings for the automotive and industrial markets, as well as IoT applications. The company’s Essentials of Edge Computing eBook (published in 2021) provides an excellent introduction to edge computing technologies and applications. Follow @nxp on Twitter for updates.
Edge computing and data centers
As data center providers, these two companies have interesting roles to play in edge computing, since data center proximity can have a significant impact on the latency of a connection.
EdgeConneX builds and operates data centers in more than 40 locations around the world. Additionally, the company builds out smaller scale solutions it calls Edge Data Centers, intended to provide low latency and high performance, and Far Edge Data Centers, which provide points of presence to connect customers who need site-specific or ultra-low latency solutions. The company also offers a network of commercial electric vehicle charging sites, as well as sites for autonomous vehicles to offload data. Follow @edgeconnex on Twitter for updates.
Equinix may be best known as a leading multi-tenant data center operator that manages more than 240 data centers located in 27 countries around the world. These sites allow customers, such as Netflix and Cisco, to leverage Equinix infrastructure. Equinix may eliminate the need for a customer to build data centers in various places. The company also offers virtual network services, precision time-as-a-service and many other connectivity options, all designed to reduce latency and enhance speed. Follow @equinix on Twitter for updates.
Edge computing equipment providers and integrators
The companies in this category tend to provide devices, connections along with the intelligence and expertise needed to make everything work together, from edge to on-premises to cloud.
From Cisco’s point of view, the “edge is anywhere that data is processed before it crosses the Wide Area Network.” The company’s edge computing offerings leverage the company’s traditional networking expertise and seek to minimize latency and network loads. Cisco IoT Solutions include sensors to monitor refrigeration, warehouses, equipment and remote facilities, as well as the Cisco IOx application environment for software development. Cisco partners with Qwilt to provide a global content delivery platform designed for modern media, such as streaming video. Follow @cisco on Twitter for updates.
Dell Technologies offers all sorts of devices designed to meet enterprise edge computing needs, including servers, storage and networking equipment. The company promotes Dell Validated Designs as “tested and proven configurations” intended for use in specific tasks and sectors, such as AI, manufacturing or retail. For example, Dell Edge Gateway and Dell Rugged Laptops are designed to gather data in harsh edge environments, while the Dell Streaming Data Platform helps organizations manage streaming data in real-time. Follow @dell_edge on Twitter for updates.
A fully-owned subsidiary of Hitachi, Hitachi Vantara has expertise in data management and storage systems, IoT and business intelligence. The company’s Lumada IoT offerings provide solutions for manufacturing, field service management, asset management and industrial data operations, among other services. The name, Lumada, is intended to suggest the ability to illuminate data. Pentaho, an open-source data integration and business intelligence platform, is part of the Lumada offering from Hitachi. Edge-to-cloud infrastructure offerings help enterprises manage infrastructure, storage and private, hybrid and multi-cloud Kubernetes. Follow @hitachivantara on Twitter for updates.
Hewlett Packard Enterprise
Hewlett Packard Enterprise offers the HPE GreenLake edge-to-cloud platform, which the company promotes as “the cloud that comes to you” and combines the company’s deep hardware expertise with cloud services and management capabilities. HPE Edgeline ruggedized systems are designed to operate in harsh environments, ranging from manufacturing facilities to mobile vehicles. Aruba Edge Services Platform offers automated network management and edge-to-cloud security services. HPE IoT solutions operate in a variety of manufacturing and industrial environments. Follow @HPE and @HPE_GreenLake on Twitter for updates.
IBM’s Edge computing focus leverages the company’s IBM Cloud and consulting expertise. IBM Edge services help companies adopt edge, IoT, AI and 5G solutions, design and adopt hybrid cloud configurations, and strategically select and deploy edge sensors, devices and networks. IBM Edge Application Manager provides autonomous management (i.e., shifted from a human administrator to software) for edge computing, intended “to help clients move data and applications seamlessly across hybrid cloud environments, from private data centers to the edge.” Follow @IBMcloud and @IBMIoT on Twitter for updates.
Edge computing software, systems and platforms
This set of edge companies tends to focus on a specific aspect of edge computing. Many are building the software tools needed to connect and manage edge devices and systems smoothly.
Alef seeks to help companies create, deploy and manage private mobile networks. The company’s set of APIs, eSims and other tools seek to reduce the complexities of private mobile networking of Citizens Band Radio Service (CBRS) enabled equipment, such as mobile devices or IoT sensors, as these devices connect to an Alef Edge Point. Alef refers to their overall offering as a mobile network-as-a-service. The company identifies uses for these private networks in sectors such as warehousing, manufacturing, higher education, healthcare and finance. Follow @alef_edge on Twitter for updates.
IOTech Systems offers software platforms for IoT development, deployment and management that include Edge Xpert, an open software platform for IoT, Edge XRT for time-critical and resource-constrained IoT systems and Edge Builder to manage edge applications and nodes. The company also seeks to simplify connecting to IoT devices via various operational technology protocols (e.g., BACnet, Bluetooth LE, MQTT, zigbee) and linking edge devices to cloud systems (e.g., AWS, Microsoft Azure, Google Cloud, IBM Watson IoT). Follow @iotechsystems on Twitter for updates.
Mimik focuses on software to allow developers to build apps for edge devices that reduce the need for data to travel to the cloud. The company’s edgeEngine seeks “to add server capability to any app or smart device” with the target systems ranging from phones to drones to IoT gateways and even robots. mimik’s FreeRTOS edge SDK is intended to transform sensors into two-way communications devices securely. Their library of microservices can help reduce development time for edge, IoT and other systems. Follow @mimiktech on Twitter for updates.
The Ori Global Edge platform seeks to simplify edge-to-cloud management. Ori Edge helps manage workloads across on-premise systems (e.g., your enterprise data centers), private cloud systems and edge sites, such as a factory or warehouse. Ori MEC, or Multi-access Edge Computing, lets telecommunications companies provide access to networks. Ori Multi-Cloud offers management orchestration services of workloads across AWS, Microsoft Azure and Google Cloud. The Ori Global Edge Marketplace will give access to edge-ready applications from several different partners. Follow @ori_edge on Twitter for updates.
A founding member of the Streaming Video Alliance, Qwilt initially focused on delivery of low-latency video, which attracted customers such as Disney Streaming and Valve. The company’s Open Edge Cloud Platform combines software edge nodes, cloud-based open APIs and a centralized management system. In May 2022, the company launched Qwilt Site Delivery to deliver last-loading websites. Qwilt has received significant funding from Cisco Investments and also has a significant strategic partnership for content delivery with Cisco. Follow @qwilt on Twitter for updates.
Acquired by Ekinops in November 2021, the main offering of SixSq is Nuvla.io, a platform that helps people deploy, manage and update edge applications. The Nuvla.io marketplace works in conjunction with the company’s NuvlaBox system, which can “turn any ARM or x86 single board computer into a smart edge device” to provide a full software-as-a-service solution for edge computing apps in container format. The Nuvla edge and container management software is open source and available on Github. Follow @sixsq on Twitter for updates.
Trilogy Networks’ FarmGrid offers farmers a solution that combines network connectivity with applications that use machine learning and modeling. Farm-to-cloud applications help with monitoring and management of row and specialty crops, as well as livestock and dairy operations. The company also is building out both ConEx, a United States-wide network, and LinX, a platform designed to simplify edge connectivity. The company’s Rural Cloud Initiative seeks to build a coalition of rural telecom providers and enhance edge computing in rural communities. Follow @trilogynetworks on Twitter for updates.
Vapor IO wants “to re-architect the internet at the edge, powering a new generation of applications — for good.” The company is building out compact, multi-tenant, modular data centers; custom fiber networking and software to manage it; software such as the Kinetic Edge Portal to manage these data centers and networks; and an open source software system, Synse, to monitor and manage workloads. These systems can be deployed in multiple sites in a city or region for high reliability, resiliency and responsiveness. Follow @vaporio on Twitter for updates.
Viso focuses on delivering a no-code/low-code platform for computer vision applications that handle tasks such as object counting, license plate recognition, eye gaze detection and face recognition, among many others. Viso Suite lets you build and deploy your vision application to edge devices, integrate and analyze connected camera video streams, as well as monitor performance. The system includes more than 55 AI models you may use with no coding needed, or you may import your own AI models. Follow @viso_ai on Twitter for updates.