Data centre infrastructure in South Africa has enjoyed a significant boost thanks to continued investments by hyperscalers and the increasing need for AI-ready infrastructure.
By Johan Robinson, EMEA AI platform leader at Red Hat
Case in point, Microsoft plans to spend R5,4-billion over the next three years to meet the growing demand for Azure services in the region.
This investment will give both public and private organisations access to cloud and AI solutions. Meanwhile, according to a 2024 Ipsos survey, AI usage is on the rise locally, with South Africans’ excitement surrounding the technology’s potential now outweighing concerns. The stage is set for AI to play a greater role in our daily lives.
That said, data centre infrastructure is a prerequisite for AI adoption. The sector enables enterprises to scale their IT infrastructure and kickstart their AI projects. Part of this involves knowing what approach they need to take when it comes to infrastructure.
AI is not a technology reserved for giant multinational companies with unlimited IT budgets. Thanks to common use cases and implementations, combined with the potential to scale those implementations, enterprises of all shapes and sizes can leverage AI to its full potential.
The role of the cloud
Central to any AI project or initiative is the level of flexibility, scalability and cost-effectiveness with which businesses can pursue and achieve them. The economics of cloud computing have enabled enterprises to pursue their own AI projects.
Before the widespread availability of cloud services and infrastructure, AI work was isolated and expensive, too process and resource-intensive for most businesses to implement and support. But the floodgates have now opened. The global demand for data centre capacity, in part thanks to the pace of adoption of advanced AI use cases by businesses, is predicted to more than triple by 2030, with hyperscalers like Microsoft Azure fuelling most of today’s demand for AI-ready data centres.
It is here that the need for businesses to take a hybrid cloud approach becomes clear. When pursuing an AI initiative, organisations need to consider both their existing computational capacity and the data they’re handling, specifically where that data is stored and processed and who can access it. Hybrid cloud environments combine on-premise and cloud resources for maximised agility.
Organisations can move data and AI workloads between environments and allocate resources where necessary. They also benefit from enhanced security and compliance when it comes to data storage and sovereignty, and they open the door for enterprises to leverage larger AI models that require significant computational capacity.
What can AI do for the common enterprise?
Even before South African businesses integrate AI into their organisations, it can help them make better use of their cloud computing resources.
AI-powered tools can help organisations optimise their cloud spending by identifying usage patterns, predicting costs, detecting anomalies or unnecessary instances, and helping teams find opportunities to further optimise their infrastructure. The same goes for AIOps, AI for IT operations, where teams can streamline operations and service management thanks to increased observability and real-time insights.
From there, it becomes a question of what you want AI to do for you. For most businesses, AI is a powerful and cost-effective tool for customer service, sales and marketing, where applications can help businesses achieve greater insight into their customer relationships, set up chatbots and automated channels for greater convenience, and develop more effective marketing strategies.
Most South African businesses will also see the value in generative AI (GenAI) programmes, which can automate content creation, especially useful for marketing purposes.
Importantly, all these use cases are feasible even for smaller, more budget-restricted enterprises. This is how AI will eventually become a cornerstone of any modern business, whether a small start-up or a giant multinational.
Overcoming challenges and scaling upwards
Integrating AI into businesses is not always as simple as downloading and installing an application. Businesses that are looking to build and maintain their own applications and solutions need to have a robust strategy in place that takes all variables into account, including common challenges.
These include having comprehensive data handling and management protocols in place, the necessary tools and integrated platforms, and the knowledge and expertise needed to design, train and deploy AI models.
With hybrid cloud infrastructure, organisations can adopt an open technology architecture that lets them scale their AI applications and workloads across multiple IT environments, enabling them to match the pace of business demand. This also lets organisations, teams and departments work together and collaborate.
Organisations also need to nurture a cross-functional, multidisciplinary team that not only includes all stakeholders, but is backed by a thorough understanding of business goals. Finally, businesses need to be selective when it comes to projects.
They need to pursue those with a high likelihood of success, which enables them to build momentum for future ones, all while learning from any potential mistakes and laying a standardised foundation for AI implementation.
Performance and consistency are critical for success with AI. With the right approach to the cloud and with platforms that offer consistent operability across all environments, businesses in South Africa can capitalise on the country’s increased data centre capacity and contribute to a fully realised, AI-powered business landscape.