AI continues to dominate the storage conversation

As more businesses look to deploy and scale AI at the heart of their operations, they are running into data and infrastructure challenges that often require new capabilities. This is particularly the case as organizations look to embrace inference at scale, using their own unique data.

Research reveals that AI is the dominant driver of storage budget growth, with the primary storage challenge around performance. Additionally, integrating the storage infrastructure with the broader data environment is a growing challenge at both a technology level and an organizational level. Only a third describe the relationships between data and infrastructure teams as 'highly collaborative.' With pressures building on organizations to deliver meaningful value from their AI investments, effective integration here is shaping up as a critical success factor.

Storage strategies increasingly span the hybrid cloud

Although the typical organization has multiple flavors of storage deployed -- such as SAN, NAS, HCI, software-defined, public cloud and others -- 'hybrid cloud storage' is now the most dominant overall deployment model. This represents a marked shift over previous studies, highlighting a growing understanding that managing storage and data across an organization's entire IT environment -- both on- and off-prem -- is increasingly important to customers.

This should be no surprise, as research also shows that, for most organizations, their data is highly distributed across the hybrid cloud. Moreover, the emergence of AI as a truly hybrid workload is raising the need for an effective data control plane that spans the hybrid cloud. Almost three-quarters of respondents (74%) said improved integration between on-premises and public cloud storage would drive substantial benefits.

Storage flexibility is more important than ever

Closely related to the above point, the relationship between the storage environment and the broader software and application infrastructure is also changing. The last two decades saw the two move closer as organizations standardized around a single virtualization stack or cloud provider. Multiple factors are calling for increased flexibility, such as:

• The need to support new or additional hypervisors.
• The need to support container architecture as well as hypervisors.
• The fast-growing need for some customers to deploy infrastructure that better supports data sovereignty requirements.

In the past, the challenge for customers was that this flexibility came at the expense of simplicity. In the age of the IT generalist and with infrastructure spanning the entire organization -- core data center, cloud and edge -- this tradeoff will no longer be tolerated.

Storage and data security are now business-critical

A further notable finding from the research is the degree to which storage systems may represent a vulnerability. More than half of respondents said they believe their on-prem primary storage was 'highly' or 'very highly' vulnerable to a cyberattack. No surprise, then, that enterprise storage buyers will prioritize enhanced security features when it comes to their next storage purchase. Particularly, features that both secure the physical storage environment and protect the data that resides on those systems.

As the threat landscape becomes ever more sophisticated, the ability to both protect the IT estate and recover quickly if, or more likely when, an attack is successful, is no longer just an IT requirement but a business requirement. However, there's less agreement among IT decision makers when it comes to precisely the best way to address these issues, highlighting the role for greater education and understanding among IT decision makers.

Procurement methods evolve as future infrastructure requirements become uncertain

IT organizations use a range of methods to procure their on-premises storage today, typically utilizing traditional Capex and IT leasing models. However, emerging workloads such as AI make future infrastructure requirements difficult to predict.

As a result, a majority of IT decision-makers are evaluating alternative procurement approaches for their next storage purchase to afford them greater flexibility amid this uncertainty. Key among these are OpEx-centric approaches, with cloud-like 'as a service' models, as well as third-party Opex offerings, under consideration. This is another area that has seen rapid development on the solutions side from storage providers in recent years; as customer requirements continue to evolve rapidly, the need for continued evolution here seems apparent.

Findings from this latest storage research study serve to underscore that the structural infrastructure shifts we have seen in recent years are accelerating. The fundamental role of enterprise storage infrastructure in storing and protecting an organization's data assets remains unchanged. Still, a barrage of growing pressures from multiple directions means the modern storage environment is being asked to do much more. That is, to be more secure, more intelligent and more flexible.

Of course, there's a final, critical curveball that IT decision-makers must also navigate: the global supply chain crisis, which is significantly raising storage and memory prices and elongating procurement cycles. These additional pressures don't negate any of the findings of this latest research. If anything, they further underscore the imperative for storage modernization. Our next storage research study will offer deeper insights into how this crisis is further impacting customers and how they are responding.

Simon Robinson is principal analyst covering infrastructure at Enterprise Strategy Group, now part of Omdia.

Enterprise Strategy Group is part of Omdia. Its analysts have business relationships with technology vendors.