As AI adoption accelerates, tech infrastructure vendors face a
strategic paradox. Despite recent positive earnings reports and
surging demand for AI infrastructure, profitability is increasingly
captured elsewhere in the value chain.
Semiconductor companies are capturing a disproportionate share
of the margins, while hyperscalers are vertically integrating by
developing custom silicon and in-house infrastructure. Meanwhile,
ODMs (original design manufacturers) are driving scale by
delivering cost-efficient, white-label AI hardware. This leaves
tech infrastructure vendors caught between high-margin chipmakers,
powerful cloud platforms, and cost-driven ODMs—and struggling
to maintain dominance in an ecosystem rapidly evolving beyond
traditional market structures.
Past disruptions in the tech infrastructure value chain
This is not the first time an industry-defining disruption has
threatened the tech infrastructure value chain. Over the past 30
years, the rise of the internet, transition to mobile-first
ecosystems, and emergence of public clouds have all reshaped market
dynamics.
Internet and digitization: Standardized
architecture in the late 1990s and the internet boom of the early
2000s sparked a surge in demand for infrastructure, driving x86 as
the de-facto standard computer architecture and rapid on-premise
enterprise IT buildouts. Industry players such as Dell and HPE
expanded aggressively and capitalized on the shift, as servers
became strategic assets in the enterprise stack. From 1995 to 2007,
Dell’s revenue grew at ~26% CAGR—its fastest-growth era
to date, according to CapitalIQ data. Infrastructure vendors
wielded significant negotiating power by serving a fragmented set
of enterprises.
Public cloud: The rise of public cloud systems
in the early 2010s marked a shift as hyperscalers like AWS,
Microsoft, and Google started designing their own infrastructure,
reducing dependence on traditional infrastructure vendors. For
example, between 2007 and 2015, Dell’s infrastructure business
achieved a lower revenue CAGR of ~8%, while margins stalled amid
weakening pricing power. As hyperscalers internalized more of their
tech stacks, infrastructure player influence in enterprise IT
declined.
Data-centric and ML workloads: From 2016 to
2022, the growing adoption of data-centric and machine learning
(ML) workloads accelerated the shift toward software-defined and
cloud-native infrastructure. Value shifted to data platforms and
orchestration layers. Both infrastructure players and hyperscalers
benefited from this growth. HPE’s server segment achieved a
CAGR of ~2%, while Dell’s Infrastructure Group posted a CAGR of
~10%. Meanwhile, the leading cloud providers grew even faster with
AWS reporting ~37% CAGR and Microsoft Azure’s ~20% CAGR.
Infrastructure vendors started to lose control of the software
stack, while Supermicro began to stand out by delivering modular AI
systems tailored to emerging enterprise needs.
AI boom marks a pivotal inflection point for infrastructure
players
The recent boom in generative AI is rapidly shifting computing
demand towards large-scale GPU clusters and vertically integrated
systems. The following key trends are combining to impact
infrastructure vendors:
Semiconductor giants expanding downstream:
Chipmakers such as NVIDIA (CUDA, DGX, AI Foundry) and AMD (ROCm)
are moving beyond silicon to full-stack platforms, including
software libraries, reference designs, and AI
frameworks—putting them in direct competition with
traditional infrastructure vendors. They are also collaborating
with AI native providers like CoreWeave (in which NVIDIA holds a
~7% equity stake as of March 2025) to extend their ecosystem around
modular AI infrastructure.
Hyperscalers are moving upstream: Cloud giants
such as AWS (Trainium, Inferentia), Microsoft (Athena), and Google
(TPU) are developing custom silicon and vertically integrating
hardware / software stacks, reducing reliance on traditional
infrastructure vendors, and consolidating control over AI
infrastructure.
ODMs are driving commoditization: Companies
such as Foxconn, Quanta, Inspur, and Supermicro are delivering
white-label, cost-efficient AI hardware at scale, pressuring
traditional infrastructure providers. Supermicro, acting as both a
vendor and ODM, is fueling its growth (78% revenue growth between
2022 and 2024) by meeting demand for ready-to-deploy, modular GenAI
servers.
AI demand is concentrated among a few large
customers: Hyperscalers and frontier AI labs like OpenAI,
Meta, and Anthropic are consuming a disproportionately high share
of AI computing, resulting in concentrated buying power and making
it harder for vendors to diversify revenue.
Rise of edge computing: The need for
low-latency AI inference in retail, industrial, and telecom
environments is accelerating demand for AI-capable edge
devices—opening new growth markets for lightweight,
energy-efficient hardware.
Focus on energy efficiency: With skyrocketing
compute demand, energy use has become a bottleneck. Infrastructure
vendors can gain an advantage by innovating in cooling, packaging,
and chip design that is optimized for AI power / performance
trade-offs.
To remain relevant, we believe infrastructure vendors must
commit to a clearly defined and well understood strategy.
What can infrastructure players do?
We see four strategic archetypes that can help infrastructure
vendors chart a successful path forward in this evolving
infrastructure value chain.
Each archetype—whether focused on innovation, customer
service, speed of execution, or cost leadership—requires
distinct choices in product design, go-to-market strategy, and
investment priorities. Mixing elements from multiple archetypes may
appear attractive but often results in weak differentiation and
organizational inefficiency.
This is a tried and tested matrix, honed across industries over
time. Infrastructure vendors can now leverage it to navigate this
disruptive moment.
Technology innovators invest heavily in
R&D and differentiated designs to maintain a competitive edge.
Tesla is a great example of a technology innovator redefining a
mature industry, through relentless investment in electric vehicle
technology, autonomous driving, and battery innovation. Another
example is Airbnb, which disrupted the hospitality industry by
leveraging digital platforms and data to create a scalable,
user-driven marketplace.
Customer service leaders leverage strong
customer relationships and offer industry-focused solutions built
on trust and long-term value. Costco built its reputation through
consistent customer value, generous return policies, and a
membership model that fosters loyalty and long-term relationships.
Similarly, American Express differentiates itself through premium
customer service, concierge offerings, and strong fraud protection,
positioning itself as a trusted partner for both consumers and
businesses.
Fast followers rapidly replicate emerging
trends and scale operations in growing markets. Their success
depends on effective go-to-market execution, leveraging broad
distribution and partnerships, and differentiating through
price-value messaging. Inditex, the parent company of Zara,
mastered fast fashion by rapidly translating runway trends into
affordable retail offerings, supported by agile supply chains and
frequent inventory refreshes.
Low-cost players compete primarily on price
and high-volume sales, maintaining a very low-cost structure. For
example, Ryanair became the European budget airline market leader
by stripping down services to essentials, maximizing aircraft
utilization, and maintaining ultra-low operating costs. In the
retail industry, Temu gained traction by offering low-priced goods
sourced directly from manufacturers, leveraging aggressive
marketing and data-driven merchandising tactics.
As the tech infrastructure market evolves under AI-driven
demands and margin pressure, many players may need to turn to
M&A to solidify their position within a chosen archetype.
Innovators may acquire specialized IP or AI infrastructure
companies to deepen their differentiation. Hewlett Packard
Enterprise (HPE)’s recent acquisition of Juniper
Networks, in a deal valued at $14 billion, will allow HPE to
strengthen its portfolio with Juniper’s AI-native networking
expertise. Palo Alto Networks’ planned acquisition of CyberArk
for $25 billion will allow Palo Alto to expand its platform into
identity security and lead the next generation of cyber
defense.
Customer service leaders can enhance their value by acquiring
service platforms or vertical-specific solutions. Cisco’s
acquisition of Splunk in March 2024 for $28 billion will allow
Cisco’s networking and security solutions to incorporate
Splunk’s data analytics capabilities to offer enhanced
visibility and insights for customers. Fast followers might use
acquisitions to quickly adopt validated technologies or expand into
adjacent segments like software-defined networking. Meanwhile,
low-cost leaders may consolidate with peers or acquire automation
and supply-chain capabilities to further drive scale and
operational efficiency. In each case, focused M&A can help
vendors fill capability gaps and defend their strategic
position.
The AI evolution introduces both disruption and opportunity for
tech infrastructure players. Those that effectively align their
product focus areas and commercial strategies can secure their
position in the value chain.
The content of this article is intended to provide a general
guide to the subject matter. Specialist advice should be sought
about your specific circumstances.