The 'on-premises as-a-service' genie is out of the bottle
While on-premises storage as a service has been around for a few years, it's having a moment. Organizations seek simplicity, and vendors are ready with offerings.
We provide market insights, research and advisory, and technical validations for tech buyers.
Published: 19 Sep 2024
The public cloud model has transformed the way many IT organizations consume infrastructure to support their digital initiatives, but "cloudifying" on-premises IT has taken more time to evolve. However, research from TechTarget's Enterprise Strategy Group, along with a slew of industry developments, suggests this is now picking up pace.
Until relatively recently, IT organizations had little choice when it came to deploying, managing and paying for their core IT infrastructure: Either buy or lease equipment and systems to run on premises, or outsource the process to a third party.
When the public cloud model emerged about 15 years ago, everything changed. Rather than buying infrastructure in multiyear cycles, IT organizations could tap into seemingly limitless amounts of compute power and storage on demand, literally at the press of a button.
On-premises as a service steps into the light
Public cloud adoption has skyrocketed, as organizations large and small have come to appreciate the model for its simplicity, access to the latest technologies and transparent, pay-as-you-go pricing. It's little surprise the cloud model took off with such gusto. The traditional model tied up both capital and labor resources for multiple years and forced enterprises to guess future hardware and related infrastructure requirements in a fast-changing world.
But although many organizations now have a "cloud-first" or even "cloud-only" posture, most applications globally still run on premises, and many will likely remain that way. However, this trend leaves scope for improvement in the way these applications run. For most organizations that still have some, most or all of their IT estate on premises, the appeal of extending the cloud model there -- creating a true hybrid environment -- remains strong. Until recently, though, it has been frustratingly difficult to achieve.
This might be changing. We are now seeing a boom in the availability of on-premises, as-a-service offerings, also referred to as storage as a service, that enable organizations to deploy infrastructure on premises, deployed and managed by a third party on a cloud-like, metered basis. Though interest in such offerings has been strong, actual adoption has taken time to build. However, there's strong evidence that the use of as-a-service offerings overall is now rapidly increasing.
Our data suggested the appetite for on-premises as a service is strong. According to our research, fewer than one in five IT organizations want to procure, manage and maintain the hardware that supports their on-premises applications themselves. Sixty percent prefer flexible consumption models, roughly half of whom would prefer their infrastructure to be managed by a third party.
Although running an on-premises environment provides organizations with high degrees of control, there's no getting away from the fact that running a large IT infrastructure is already complex. For most, it's getting more complex.
According to Enterprise Strategy Group research, nine in 10 organizations said their IT environment has become more complex in the last two years. For many, the sheer number of initiatives that are underway compounds the challenge. Enterprises undergo digital transformation, security initiatives, application replatforming to cloud-native architectures, and exploration and adoption of emerging technologies such as AI.
IT decision-makers are attracted to on-premises as-a-service approaches because they deliver a cloud-like experience for their on-premises workload.
At the same time, many infrastructure professionals report they are already struggling to keep up. Eighty percent of respondents to a recent study noted they had recently taken on additional responsibilities to support their organization's digital transformation or were under pressure to do so.
IT decision-makers are attracted to on-premises as-a-service approaches because they deliver a cloud-like experience for their on-premises workloads. Notably, this isn't just because they provide access to a pay-as-you-go consumption model; they also provide faster access to the latest technologies, which help drive innovation.
This aspect is of particular interest when it comes to emerging areas, such as AI. The promise of AI might be high, but so are the up-front costs. Many organizations face difficult decisions around when, and how much, to invest, especially when the market is so nascent and the overall environment uncertain.
Storage offerings aplenty
Many organizations store much of their most important data on premises. For such firms, there's a strong preference to build an AI environment inside their own data centers. Given these factors, an on-premises as-a-service model for AI has clear appeal. An organization can deploy -- and pay for -- an integrated AI infrastructure stack that starts small, scales over time and can potentially be refreshed to run the latest GPUs.
The other barometer of market interest is on the supply side. Here, we have seen a marked uptick in activity, as almost all infrastructure providers now offer on-premises as-a-service options. Many are investing considerably in expanding and deepening those offerings. Recent examples include the following:
Dell. Initially launched in 2021 with a focus on metered data storage services, Dell's Apex initiative has since expanded to cover a broad range of cloud-based services and simplify customers' cloud experience. Apex subscriptions are now available across Dell server, storage, data protection, cloud, converged and hyperconverged infrastructure. Recent areas of development for Apex include AIOps and support for Red Hat OpenShift Virtualization.
Hewlett Packard Enterprise. HPE was one of the pioneers of the on-premises as-a-service model with its GreenLake initiative, first launched in 2018. GreenLake now spans a broad range of HPE capabilities that help customers across compute, storage, AI and hybrid cloud management initiatives. HPE continues to invest in GreenLake as a core part of its differentiation. The company's pending acquisition of Juniper looks set to extend GreenLake's as-a-service credentials further into the networking realm.
Hitachi Vantara. Noting growing customer demand for simplified operations and consumption models that match business needs, the storage specialist is placing a fresh emphasis on its EverFlex IaaS offering. EverFlex includes a comprehensive as-a-service catalog spanning storage, data protection, compute, converged infrastructure and FinOps, all under a new SaaS-based management portal. The platform is supported by a range of consulting services focused on elements such as containerization, AI workload enablement, private cloud and specific applications such as Oracle and SAP.
IBM. "Big Blue" offers a range of consumption-based models for its storage capabilities. Earlier this year, the company announced IBM Storage Assurance. The new subscription service provides FlashSystem customers with hardware and software upgrades along with premium support, over a four- or eight-year term, with flat and transparent pricing.
Lenovo. The infrastructure specialist continues to double down on its TruScale initiative. It extends its roots in storage and compute to encompass "everything-as-a-service" that spans client devices, AI, data protection, sustainability and more. A notable recent addition is Lenovo TruScale GPU as a service, offering metered Nvidia GPU resources and orchestration. It provides customers with immediate and scalable access to AI and HPC technology.
NetApp. Though NetApp's Keystone storage-as-a-service offering was first released in 2019, it's now garnering more attention as interest increases. NetApp said Keystone revenues increased 60% year over year in its most recent quarter. NetApp points to Keystone's ability to help customers reduce risk in an uncertain environment. That's not just financial risk either; by embedding its proactive Autonomous Ransomware Protection capabilities into Keystone SaaS, it's able to help customers address security risks as well.
Pure Storage. The all-flash storage specialist helped transform painful legacy issues in storage, such as disruptive, forklift upgrades, with its Evergreen program. Pure continues to challenge the status quo in myriad ways, with a growing range of as-a-service and subscription capabilities that span both on-premises and cloud-based environments. Subscription services, which are delivered entirely through channel partners, already account for about half of Pure's overall revenues.
IT decision-makers have never had as much choice when it comes to consuming on-premises infrastructure, which, increasingly, works in better harmony with public cloud environments.
The on-premises as-a-service model still isn't going to be for everyone. Some organizations prefer to stick with the traditional Capex model, especially for well-established, stable applications that they understand well.
However, for emerging applications with uncertain long-term requirements such as AI and for more unpredictable or spiky applications, the on-premises as-a-service model is increasingly compelling. It should help encourage more organizations to embrace the future in a way that minimizes risk. Now that the on-premises as-a-service genie is out of the bottle, it doesn't look like it's going back in.
Simon Robinson is principal analyst covering infrastructure at TechTarget's Enterprise Strategy Group.
Enterprise Strategy Group is a division of TechTarget. Its analysts have business relationships with technology vendors.
