Even within a technology industry known for hyperbole, the growth of the artificial intelligence (AI) market is incredible. Last summer, a Bloomberg Intelligence report proposed that the AI market would reach $1.3 trillion over the next 10 years. A significant increasse from a market size of ‘just’ $40 billion in 2022. Most recently, June 2024 saw AI giant NVIDIA hit a value of $3tn, eclipsing Apple. The iPhone maker immediately responded with the launch of Apple Intelligence AI. And the race between major tech firms shows no sign of slowing.
A critical part of this growth is that AI has rapidly evolved beyond being the exclusive province of large corporations. The ease of use enabled by AI interfaces has led to businesses of all sizes embracing the technology.
In particular, high-growth, medium-sized businesses (MSBs) recognise the potential of AI. In the UK, there are approximately 35,900 such MSBs. For these organisations, the possibility to automate tasks and accelerate decision-making is a huge source of competitive advantage. However, successfully embracing AI requires a strong foundation of digital infrastructure that such organisations often overlook.
The right framework
An AI-ready digital infrastructure can be broken into four key areas:
High-performance networking: Inference AI applications need networks that reduce latency to the edge for in-flight analytics and real time data processing. Training AIs need high, 100 Gbps bandwidth to the data centres or cloud where large training datasets are stored. This connectivity must be highly reliable, with multiple connections and bandwidth across resilient, fast networks, and secure data transfer protocols.
Secure data storage: Artificial intelligence lives and dies by the data it is ‘trained’ on. If an MSB is set to embrace AI, it must have a secure, scalable data storage solution to house both the structured and unstructured data that is used for training and running AI models.
Data management and governance: Extracting value from the data used by AI, requires effective data management practices. And it is not just a commercial imperative. MSBs need robust data governance frameworks in place to ensure compliance regulations. Establishing secure pipelines to automate the collection, organisation, and preparation of data for AI is crucial.
Regional edge: While cloud computing offers immense power for some AI use cases, it can introduce latency issues for applications that require real-time decision-making. Regional edge computing puts processing power closer to the source of data. This reduces latency and enables faster processing of time-sensitive data. This has already shown its value in applications such as predictive maintenance or real-time video analytics.
Diving deeper – AI and the regional edge
MSBs typically operate in geographically dispersed locations, dealing with real-time or near-real-time data streams in order to serve markets faster. In these contexts, regional edge computing offers significant advantages.
• Reduced latency: By processing data closer to its source, regional edge computing minimizes the time it takes for data to travel between collation and processing. This is crucial for applications requiring immediate insights and decisions, such as real-time anomaly detection for fraud, or optimising dynamic supply chain logistics.
• Improved bandwidth efficiency: Edge computing reduces the amount of data that needs to be sent back to centralised facilities, freeing up valuable bandwidth and lowering network costs.
• Enhanced security: Sensitive data can be processed at the edge before being sent to the cloud or elsewhere. This reduces the security risks associated with data transmission over long distances.
With these sorts of benefits, it is little surprise that a recent report found that 77% of executives say their technology architecture is either very critical or critical to the overall success of their organisation.
When it comes to AI, regional edge opens a pathway to cost-effective deployment. In the recent 2024 Trends in Datacenter Services & Infrastructure report from S&P Global Market Intelligence, the analysts note: “…the rise of AI inference workloads that may also have latency and data location requirements could further drive edge deployments.”
The same report also notes that “…use cases may vary enormously, so it may be hard for vendors to gain scale in an atomized market. The ecosystem of vendors, operators, financing and network providers at the edge is evolving rapidly.”
Partnering to drive AI
The idiosyncratic demands of MSBs looking to embrace AI will be best served by regional digital infrastructure providers. These providers can partner with MSBs to address new use cases in the face of profound industry challenges.
Given the fragmented market, and technological demands, building and managing a complex foundation for AI is daunting for MSBs. This is especially true in the face of skills shortages throughout AI and infrastructure alike.
As far back as 2021, the UK Government identified that nearly half (49%) of UK firms had been affected by a lack of technical AI skills, and almost a third (32%) had been similarly impacted by a lack of non-technical capabilities.
Similarly, 2023 figures from the Uptime Institute showed that more than half (53%) of UK data centre operators report having difficulties in finding new talent, up from 38% in 2018.
In the face of this, the expertise and guidance of experienced technology partners carries major benefits. It means a faster implementation, optimised costs and compliance with demanding data privacy regulations.
By leveraging a digital infrastructure partner that can combine high-performance networking, secure data storage, cloud options, and the emerging power of regional edge computing, MSBs can approach AI methodically and with minimal disruption to ongoing business, whilst navigating the opportunities that AI will undoubtedly bring.
- Digital Strategy
