A recent Forrester study on cloud infrastructure in 2023 reveals that 79% of cloud decision-makers in over 1,300 companies are implementing private clouds. Additionally, IDC predicts that global spending on dedicated private cloud services, including hosted private clouds, will reach 20 billion euros in 2024 and double by at least 2027.
Global spending on enterprise private cloud infrastructure, including hardware, software, and support services, will amount to $51.8 billion in 2024 and grow to over 65 billion euros in 2027, according to IDC. However, public cloud providers like AWS, Microsoft, IBM, Oracle, and Google will continue to dominate the market, with estimated revenues of over 800 billion euros in 2024.
The surge in artificial intelligence is driving the resurgence of private clouds, which were previously seen as little more than a solution for purchasing hardware and placing it in a data center. The popularity of private clouds dwindled due to the extensive functionalities offered by public cloud providers, far surpassing the capabilities of open-source private cloud systems or those offered by enterprise hardware manufacturers.
With the increasing prevalence and complexity of AI workloads, many organizations are reevaluating their cloud strategies. The current trend among enterprise architects suggests a transition to a hybrid cloud infrastructure.
A key factor driving this trend is the need to control the rising costs associated with cloud and AI technologies. Public cloud providers are proving to be more expensive than their regional counterparts, and this reality is finally sinking in with CIOs and CFOs who find no justification in paying costs that are 2 to 5 times higher than private cloud solutions.
Key platforms in this transition
Private cloud platforms like Dell and HPE (now equipped with generative AI support) are playing a crucial role, but not without considering European infrastructure providers like Stackscale, which offer tailored private cloud solutions with GPUs for AI. Many other solutions in the market are also claiming their space as private clouds. This “private cloud wash” was common in the early days of cloud computing but has resurfaced as the market matures.
These platforms provide the computational power, including GPUs, and the flexibility needed to handle AI workloads. They also maintain strict control over data privacy and security, although often this security is more of a perception than a reality. In many cases, public cloud providers offer better security due to their greater investments in proprietary solutions.
The rise of AI has increased concerns about data security, specifically the risk of corporate private data being “siphoned off” into public AI models without permission. While a contract may state otherwise, in public solutions, anything is possible, making it a recurring concern. The possibility of public cloud providers inadvertently accessing corporate data and using it to train their AI models would be a scandal of epic proportions. However, many companies find private cloud appealing as it allows them to retain sensitive data in a controlled, exclusive environment with the ability to know the real location, even audit it and visit regularly.
Challenges and emerging solutions
Despite its advantages, private clouds are not without challenges. Specialized hardware is required for large-scale AI operations, such as servers with GPUs, which can be costly and require extensive energy and cooling systems. Companies still do not fully understand the new updates and costs this will entail, which in many cases will be more extensive than running these workloads in public cloud providers.
However, solutions are emerging, such as building private clouds within data centers, like those provided by Equinix, Data4, Interxion, or the Spanish Grupo Aire through OasixCloud. These centers are specifically equipped to handle these infrastructure requirements and are considered better options than building everything internally. Eventually, companies will need to move out of the data center business, letting public cloud providers, hosting, and managed services offer better options.
So, is private cloud a good option for companies? It has always been on the table as a real alternative to public cloud, with effective cost control and the certainty of exclusive computing. It has its uses, and if it is more cost-effective or can deliver more value to the business, it should be utilized, with or without AI.
Private cloud will continue to advance, gaining ground on public cloud, as more and more companies reconsider entrusting their core infrastructure to national or European providers, not putting all their cards in American companies. The future lies in investing in European companies to regain ground and not fall behind.