It is fair to say the predictions for edge growth have yet to materialize fully. However, like an iceberg, much of the action is happening under the surface.
The Edge is unlike the Cloud, where the applications already existed in the data center and it was more of an infrastructure lift and shift play. At the edge, technological advancements are spurring an entirely new class of applications that require edge processing while also enabling the required infrastructure to deliver those applications.
Entering 2024, we see edge deployments moving to scale while remaining under the radar to the casual observer. Growth at the edge is a confluence of numerous technologies merging together – such as hybrid and multi-cloud, 5G infrastructure, DevOps automation, and AI – each of which are advancing and improving. All together, these are allowing for larger and more integrated deployments of edge solutions.
#1: Building Up 5G Telco Networks
Standalone 5G will roll our nationwide finally bringing the full suite of 5G capabilities. Relative to the potential what we have today is 5G-lite.
We will see ongoing expansion in the deployment of 5G Fixed Wireless Access (FWA). This is enabling telecom operators to expand outside their fixed line footprint and offer broadband nationally including underserved rural areas, increasing choice and competition. FWA will also be used to serve small businesses and enterprises. FWA can be used to rapidly add new connections and also provides a failover to the fixed network.
5G deployments will drive more edge applications moving forward. With the advent of 5G RedCap, there will be an uptick in IoT devices and applications. Edge-focused Private 5G and Neutral Hosts are being deployed in venues like sports stadiums, convention centers, train stations, metros, and airports. Leading candidates for industrial applications are in ports, harbors, mines, and smart factories. While many consumers may associate 5G with cell phone usage, the business use of 5G is exploding for businesses due to support for low latency connectivity, massive IoT and private networks. Network slicing and hybrid public-private networks enables enterprise SLA (Service Level Assurance) and multisite-site privacy and security.
All of these infrastructure improvements at the telco side to drive improvements for 5G are being met with hardware upgrades inside enterprises as well.
#2: Edge Expansion at the Enterprise
Following the recent trend of moving computing to the cloud, is the next stage or trying to move computing to the edge. But a challenge with the edge is managing the wide distribution of devices and applications. To address this, we see a rise in containers for edge deployments, with examples being Amazon’s Elastic Kubernetes Service and more tailored edge tools like Red Hat Microshift with its Ansible integration and Wind River Studio that can run on smaller single-node devices. This enables full DevOps automation for edge compute, while allowing for seamless scaling when more resources are needed.
Similarly, hyperscalers have all announced and are currently deploying edge cloud solutions. These enable users to leverage the same tools and skills used in cloud deployments for edge applications. Examples include AWS Outpost, Azure Edge, and Google Distributed Cloud Edge. This movement towards data collection and processing in a distributed fashion results in lower latencies where needed, less network traffic and faster decision making. As cloud service providers make their services available in a wide variety of environments, enterprises can keep control of their data, while taking advantage of the services that cloud providers can offer.
As new AI driven capabilities emerge and the power of processing across the board is rapidly improving at all locations – on premise, in the cloud, and at the edge – we’ll see further transformation in this space in the coming year.
#3: Advancements in AI Contribute
Across industries, especially retail; automation and Digital Transformation are spurring point-of-use edge compute, with a rise in AI assisted video, analytics, and digital signage. One such example is automated ordering kiosks for shopping checkout, which are rapidly expanding in various verticals.
This industry shift is being driven by new AI processing hardware that offers higher compute at a better price to performance ratio. GPUs, which are closely associated with AI, will be available in lower cost and less power-hungry configurations, which allow for deployment outside of environmentally controlled data centers. In addition, advances in the performance of AI directly on CPUs is enabling AI, both training and inference at the edge.
As computer performance with increased power efficiency keeps advancing, AI and Video’s higher processing needs are omnipresent. Downstream video is well over half of all internet traffic. Enabled by lower cost cameras and edge AI inferencing, upstream video is becoming ubiquitous across all application verticals and in smart cities. With the combination of video cameras and AI, edge processing will be in higher demand, controlled and managed by central systems. License plate readers, automobile identification, and other scenarios will require additional processing at the edge, in smaller packaging. AI enhanced edge processing of video will increase retail and physical security.
#4: Emerging Edge Applications
As lightly discussed in the sections above, all of these trends together are driving numerous new 5G and edge applications for several industries.
Emerging industries like logistics and robotics, requiring mobility drive 5G adoption, will also benefit from enhanced 5G Positioning. Autonomous vehicles, industrial drones, and eVTOL planes all have extensive edge compute linked to the cloud. Modern cars, especially EVs, are mobile computers remotely monitored, managed, and updated. The latest automobiles need to communicate over a network for software updates, and moving forward will need to be able to communicate with other autos, which will help to avoid collisions and lead to safer roads. Reliability of these communications will become an absolute need, as spotty Wi-Fi coverage will not be tolerated when driving on either city streets or on highways.
Looking to VR, Apple’s mid-2023 VisionPro headset launch may be the one to mainstream the metaverse. Combined with generative AI, we can foresee the rise of multimodal LLMs, which use a combination of text, speech, and images in virtual worlds, whether VR, XR, or AR-driven, to create new classes of assisted-operator professional applications and enhancements to consumer gaming and navigation guides. Imagine an operator, say a machine mechanic looking at a part which is automatically recognized, speaking to a LLM as a prompt and getting instructions via video response overlaid on the work area.
#5: Continued Digital Transformation
Across the board, businesses are adopting new technologies at a faster pace than ever before. Though the evolving state of the worldwide economy will have a significant impact on the hardware market due to global supply chains, the trend of digital transformation that started back in 2020 will still continue strong though next year. Especially as 5G, edge and AI capabilities advance further, businesses are looking to adopt advanced automation capabilities to assist with improving operational efficiencies, supporting overwhelmed staff, and handling escalating high volumes of data and traffic. Though it is unlikely that 2024 will be the year of 5G showcasing its full promise to the public, behind the scenes it will continue to shape the IT and technology industry.
link