Commentary: Dell, The Perennial Box Shifter

Watching Dell Technologies World 2025 remotely, I was surprised by the lack of products and solutions news announcements from the event. Once again, the company focused on AI hardware, with only a minor nod to server virtualisation and storage. Back in 2015/2016 when Dell acquired EMC, the opportunity existed to transform the company into an infrastructure powerhouse. So why is the focus, once again, on boxes?

Background

Michael Dell founded Dell Technologies, initially as PC’s Limited, back in 1984 from his college dorm room while a student at the University of Texas. As his book “Play Nice but Win” explains, the Dell Computer Corporation, as it became known, initially assembled PC systems from commodity components, eventually developing custom designs as the market expanded.

During the 2000s, it appeared that Dell (the company) was serious about expanding into the enterprise market, rebranding as “Dell, Inc.” to separate itself from being seen as a consumer-only brand. Dell announced its intention to buy storage vendor Compellent in December 2010, completing the deal the following February.

Perot Systems was acquired by Dell in 2009, adding a substantial services division to the company. As this 2011 news article highlights, the then-CEO Kevin Rollins mooted acquiring EMC as early as 2002, but Michael Dell rejected the idea, preferring to focus on consumer technologies. At the start of the decade, the concept of serving enterprise customers was considered risky; by the end of that period, Dell had acquired services and an expanded portfolio of products to go after enterprise customers.

EMC

The acquisition of EMC arguably pivoted Dell, Inc. into the enterprise market in a way that couldn’t have been imagined at the time. As we reported in October 2015, the purchase price of $67 billion was one of the biggest IT deals in history, leaving Dell with around $50 billion in debt.

However, Dell’s existing business, which earned around 33% of income from the enterprise, ran at a 22% gross margin, compared to EMC-II (EMC Enterprise) at 55% and VMware at 88%.

The deal looked to be a game-changer for the industry and Dell itself, propelling the company into leadership positions in enterprise storage and data protection (the former of which is still maintained today) with much better financials than Dell achieved prior to the acquisition.

However, $50 billion is a lot of debt to service. Over time, the less valuable pieces of EMC were sold off. As this article we wrote in October 2015 explains, 96% of EMC revenue came from just two divisions, EMC-II (selling storage) and VMware. By 2020, the concept of the EMC “Federation” was over when RSA was sold off to private equity.

VMware

Outside of the infrastructure products, the obvious prize for Dell from the EMC acquisition was VMware. VMware dominated x86 server virtualisation from the start of the new millennium and 25 years later is still the market leader.

Back in 2019, we questioned whether VMware would become the tip of the spear, leading a newly public Dell Technologies. The logic at the time was simple; VMware effectively owned the Private Cloud market while the margins on hardware continued to look challenging.

However, in 2021, Dell Technologies gifted VMware to its shareholders in a spinoff that mainly benefited Silver Lake Partners and Michael Dell personally. Just six months later, Broadcom announced plans to acquire VMware for $61 billion, providing Dell and Silver Lake a handsome payoff. Post acquisition, Broadcom has “sweated the asset”, turning VMware into a cash generation engine – something Dell itself could have done, but decided not to.

Back to Basics

What has Dell Technologies achieved in the last decade since EMC was acquired? Obviously, VMware and other divisions have been divested. In 2016 the software and services divisions were sold off to private equity and NTT Data International respectively.

In 2019, we discussed how Dell would rationalise the large range of storage solutions in its portfolio, representing an excellent opportunity to combine some interesting technologies. In the end, PowerStore emerged as a Clariion/VNX clone, with the remaining products eventually being discontinued. This was a pretty lacklustre effort, showing little imagination to exploit the intellectual property from a plethora of storage products available.

In the storage division, we have seen only incremental improvements in recent years, including just two hardware iterations of PowerStore, which now lags significantly behind its competitors. In contrast, vendors such as Pure Storage and NetApp are widening their portfolio of products and gaining market share as a result.

AI

The rise of AI hype has proved to be a boon for all the infrastructure vendors as customers rush to install expensive GPU-based servers. In addition, Intel is attempting to push the concept of the AI PC, an uprated device that uses embedded GPU technology to deliver many of the capabilities of large-language models (LLMs).

Dell has doubled down on AI, pushing hard on products such as the PowerEdge XE9712 and XE9680, rack-mounted servers stuffed with NVIDIA GPUs. At DTW 2025, Dell announced air and liquid-cooled servers, including the PowerEdge XE9780(L) and XE9785(L), supporting up to 192 NVIDIA Blackwell Ultra GPUs. These products form what Dell likes to call the “AI Factory”, a set of products and services entirely focused on delivering infrastructure for AI workloads.

Boxes

Does it matter that Dell has returned to its roots, focusing on selling servers rather than entire solutions? The latest Dell financial figures (annual data) show us that the gross margin for the business as a whole sits at 23.7%. If we look at the data for the past two years (available here), we can see the following.

CSG revenue declined 16% in FY2024 compared to FY2023. Revenue declined again by 1% in FY2025 compared to FY2024. Margins were better in FY2024 at 7.6% but declined in FY2025 to 6.1%. The laptop/desktop business is a competitive one, with many players, including several vendors looking to push Arm-based laptops. The AI PC has failed to generate significant spending, and implicitly, all client devices will eventually be AI PCs by default, as Intel rotates processors towards the inclusion of NPUs (Neural Processing Unit, see this post for details).

ISG revenue has fared slightly better. Revenue for FY2024 decreased 12% compared to FY2023, with Servers & Networking down 14% and Storage down 9%. In FY2025, ISG revenue increased by 29%, driven by (AI) server sales at 54%. Storage revenue remained flat at a 1% increase. Operating income in FY2024 was 12.6%, improving slightly to 12.8% in FY2025.

The sale of AI-related hardware has been the only growth area over the last two years. However, these products (as Dell has admitted previously) come with lower gross margin, resulting in a need to optimise expenses. As a result, Dell has “let go” 20% of its workforce over the last two years, as an attempt to reduce costs and stabilise operating income. Bear in mind also that at some point, the AI boom will be over. What happens to server sales then?

Competition

Dell has always been a box-shifting company. It’s in the ethos of PC’s Limited, founded over forty years ago. However, the IT landscape today is very different from that of the early 1980s. At that time, IBM and the mainframe dominated. The x86 architecture was in its infancy and consigned purely to the desktop.

Although x86 servers have come to dominate the data centre, there are challengers in the form of Arm and RISC-V. Most notable though, is the shift to the public cloud. AWS, for example, reported its most recent quarterly revenue of $29.3 billion, up 17% year-on-year. By comparison, in the quarter in which the EMC acquisition was announced, AWS made “just” $2.4 billion in revenue.

AWS makes more money now than Dell, with an operating income of approximately 40%. The AWS business continues to grow in double-digit terms, while Dell has hovered around the $90-100 billion mark for the past six years. This is just one of many public cloud platforms, all competing to steal away on-premises business.

Why are the cloud platforms gaining more traction than on-premises? The most obvious answer is investment in solutions and services rather than hardware. AWS aggressively cuts costs and develops Amazon-owned IP. The clearest example is the move to Arm-based virtual instances, a strategy also followed by Microsoft Azure, Google Cloud and OCI.

In contrast, Dell continues to develop technologies (such as its storage products) that are thirty-plus years old. The market has moved on, with vendors such as VAST Data, Weka and Pure Storage all reinventing the storage market. We are seeing similar trends in the client segment and in servers.

The Architect’s View®

Undoubtedly Dell Technologies is a massive player in the technology market and will continue to be one for many years. However, market size is no guarantee of success, as we’ve seen with the dramatic decline of Intel Corporation, in parallel with the rise of Arm Holdings PLC.

The market for on-premises IT hardware is shrinking in absolute and relative terms, as businesses place new workloads into the public cloud. Dell has no cloud strategy, as the business is focused on selling on-premises hardware to its customers (albeit with a few exceptions, such as the partnership with xAI). This is a problem for future growth and long-term sustainability.

We don’t see much changing over the next decade, particularly while the current leadership remains in place. There will continue to be a place for on-premises hardware sales, but we anticipate products from Dell and other similar vendors will decline in competitiveness, shrinking the market and pushing more workloads to the public cloud.

The answer is to return to innovation, offering customers solutions rather than products. Is it too late to reverse direction or is the die cast?