Processing video is a key use for edge computing
In our analysis and sizing of the edge market, STL Partners found that processing video will be a strong driver of edge capacity and revenues. This is because a huge quantity of visual data is captured each day through many different processes. The majority of the information captured is straightforward (such as “how busy is this road?”), therefore it is highly inefficient for the whole data stream to be sent to the core of the network. It is much better to process it near to the point of origin and save the costs, energy and time of sending it back and forth. Hence “Video Analytics” is a key use for edge computing.
Enter your details below to request an extract of the report
The edge market is evolving rapidly
Edge computing is an exciting opportunity. The market is evolving rapidly, and although still fairly nascent today, is expected to scale significantly over the next 2-3 years. STL partners has estimated that the total edge computing addressable market was worth $10bn in 2020, and that this will grow to $534bn in 2030. This is driven by the increasing number of connected devices, and the rising adoption of IoT, Industry 4.0 and digital transformation solutions. While cloud adoption continues to grow in parallel, there are cases where the increasingly stringent connectivity demands of new and advanced use cases cannot be met by cloud or central data centres, or where sending data to the cloud is too costly. Edge answers this problem, and offers an alternative option with lower latency, reduced backhaul and greater reliability. For the many enterprises who are adopting a hybrid and multi-cloud strategy – strategically distributing their data across different clouds and locations – running workloads at the edge is a natural next step.
Developments in the technologies enabling edge computing are also contributing to market growth. For example, the increased agility of virtualised and 5G networks enables the migration of workloads from the cloud to the edge. Compute is also developing, becoming more lightweight, efficient, and powerful. These more capable devices can run workloads and perform operations that were not previously possible at the edge.
Defining different types of edge
Edge computing brings processing capabilities closer to the end user or end-device. The compute infrastructure is therefore more distributed, and typically at smaller sites. This differs from traditional on-premise compute (which is monolithic or based on proprietary hardware) because it utilises the flexibility and openness of cloud native infrastructure, i.e. highly scalable Kubernetes clusters.
The location of the edge may be defined as anywhere between an end device, and a point on the periphery of the core network. We have outlined the key types of edge computing and where they are located in the figure below.
The types of edge computing
It should be noted that although moving compute to the edge can be considered an alternative to cloud, edge computing also complements cloud computing and drives adoption, since data that is processed or filtered at the edge can ultimately be sent to the cloud for longer term storage or collation and analysis.
Telcos must identify which area of the edge market to focus on
For operators looking to move beyond connectivity and offer vertical solutions, edge is an opportunity to differentiate by incorporating their edge capabilities into solutions. If successful, this could result in significant revenue generation, since the applications and platforms layer is where most of the revenue from edge resides. In fact, by 2030, 70% of the addressable revenue for edge will come from the application, with only 9% in the pure connectivity. The remaining 21% represents the value of hardware, edge infrastructure and platforms, integration, and managed services.
Realistically, operators will not have the resource and management bandwidth to develop solutions for several use cases and verticals. They must therefore focus on key customers in one or two segments, understand their particular business needs, and deliver that value in concert with specific partners in their ecosystem. As it relates to MEC, most operators are selecting the key partners for each of the services they offer – broadcast video, immersive AR/VR experiences, crowd analytics, gaming etc.
When selecting the best area to focus on, telcos should weigh up the attractiveness of the market (including the size of the opportunity, how mature the opportunity is, and the need for edge) against their ability to compete.
Value of edge use cases (by size of total addressable market by 2030)
We assessed the market attractiveness of the top use cases that are expected to drive adoption of edge over the coming years, some of which are shown in the figure above. This revealed that the use cases that represent the largest opportunities in 2030 include edge CDN, cloud gaming, connected car driver assistance and video analytics. Of these, video analytics is the most mature opportunity, therefore represents a highly attractive proposition for CSPs.
Table of Contents
- Executive Summary
- Processing video is a key use for edge computing
- The edge market is evolving rapidly
- Defining different types of edge
- Telcos must identify which area of the edge market to focus on
- Video analytics is a large and growing market
- The market for edge-enabled video analytics will be worth $75bn by 2030
- Edge computing changes the game and plays to operator strengths
- What is the role of 5G?
- Security is the largest growth area and operators have skills and assets in this
- Video analytics for security will increasingly rely on the network edge
- There is empirical evidence from early movers that telcos can be successful in this space
- What are telcos doing today?
- Telcos can front end-to-end video analytics solutions
- It is important to maintain openness
- Conquering the video analytics opportunity will open doors for telcos
Enter your details below to request an extract of the report