The Hyperscalers

Introduction

Many people hear the term “hyperscalers”, yet few truly understand who these

companies are or the magnitude of their impact. Over the past year, I’ve had the

opportunity to engage directly with several of them, gaining firsthand insight into their

scale, their decision-making dynamics, and the sheer influence they now hold over the

digital world.

When I refer to “our world,” I mean both the rapidly evolving domain of AI and the datacenter

industry that underpins it. Data centers have become the critical backbone

enabling the AI revolution to unfold.

This article builds on an excellent and timely report published by DC Byte, a global

leader in data-center market intelligence. I’ve expanded upon it with my own field

insights, practical experience, and observations relevant to the Israeli market.

What Makes a Hyperscaler - and Why That Term Matters

Hyperscalers are the dominant cloud and technology providers: AWS, Google,

Microsoft, Meta, Oracle, Apple, alongside a new generation of AI-focused cloud

providers such as CoreWeave, Lambda, Together.ai, Nebius, and Crusoe.

The term hyperscaler reflects their ability to operate at unprecedented scale:

• Scale at the gigawatt level - single campuses reaching 200–1,000 MW or more.
• Exceptional growth rates - expanding far faster than traditional players.
• Aggressive deployment timelines - they cannot wait five years for capacity;
• they need it immediately.
• Massive power requirements - a single AI cluster may consume 40–100 MW.

According to DC Byte, hyperscalers account for roughly 70% of new global demand for

data-center capacity.

In practical terms, they influence where new infrastructure will be built, which markets

will receive power allocations, and which will be left behind.

AI as the Driving Force Behind Hyperscale Expansion

One of the most important insights in the report is that today, the limiting factor for AI

growth is not GPU availability - it is infrastructure.

The world is not yet prepared for:

• Thousands of GPUs per cluster
• Power densities in the hundreds of megawatts
• Liquid cooling and immersion cooling
• Long utility lead times - often 3–7 years for new power connections

In short: Without AI-ready data centers, the AI revolution cannot scale.

Three Global Forces Shaping the Hyperscale Market

1. Power: The New Competitive Frontier

DC Byte puts it plainly:

“Power access is now the defining factor of competitiveness.”

Worldwide, power availability has become the most significant constraint:

• Northern Virginia: 5–7 years for a new power connection
• Singapore: Strict caps on new developments
• Amsterdam, Frankfurt, Dublin: Near full saturation

And in Israel:

• The national grid is already heavily loaded
• High-voltage connections take considerable time
• No unified national data-center strategy yet exists
• Government bodies are only now exploring incentives to guide development

toward underserved regions (north and south)

What is clear is this:

Whoever can deliver 40–80 MW of available power in Israel will attract the next wave of

AI infrastructure providers. The industry is consolidating - large, contiguous campuses

are becoming the preferred model.

2. Regulation: Markets That Accelerate vs. Markets That Stall

Around the world, regulatory environments are diverging:

Markets tightening regulation:

• Germany – EnEfG mandates energy efficiency and waste-heat reuse, turning thermal by-product into useful municipal heating.
• Singapore – Moratorium on large new data-center capacity.
• Netherlands – Restrictions on mega-scale developments.
• Virginia – Stricter environmental requirements and limited power allocations.

Markets accelerating investment:

• Malaysia – Became a regional hyperscale hub within two years.
• Thailand – Streamlined planning processes.
• Spain and Poland – Rapid expansion enabled by available land and power.

Israel’s position

Israel currently sits in the middle. There is:

• No national DC strategy
• No “fast-track” planning mechanism
• Significant bureaucratic friction

However, there is increasing openness among policymakers to remove bottlenecks, and

industry stakeholders - including us - are actively contributing insights from the field.

A national initiative could fundamentally shift Israel’s trajectory. Positioned along major

submarine cable routes between Europe and the Middle East, Israel could evolve into a

regional digital hub - but only with:

• Planning reform
• Dedicated high-voltage infrastructure
• Realistic assumptions about AI-era power demand

3. Speed: In the Hyperscale World, Slow Markets Get Left Behind

Hyperscalers no longer operate on traditional construction timelines. Their question is

not “When can you deliver?”

but rather: “Can you deliver on time — and if not, we will build elsewhere.”

DC Byte highlights a striking trend: 80–90% of capacity in leading markets is pre-leased

before construction even begins.

Global trends include:

• Modular and prefabricated builds
• Repeatable “capsule-based” design
• Direct coordination with power utilities

As for Israel:

We excel at building quickly, but the true bottleneck lies in pre-construction stages:

• Planning and zoning
• Environmental approvals
• Land allocation
• Power-grid connection

If these processes are streamlined, Israel can become highly competitive in attracting

cloud and AI investments.

The Global Race — Where Is the Industry Heading?

United States - Still Dominant, but Reaching Saturation

Virginia remains the global leader, but capacity and grid limitations are shifting new

development toward Georgia, North Carolina, Texas, and Alabama, where land,

regulation, and power conditions are far more favorable.

Europe - The Decline of the Traditional D-FLAP Core

Frankfurt, London, Amsterdam, Paris, Dublin — all are saturated.

Growth is now moving to Madrid, Milan, Warsaw, and the Nordics.

Asia-Pacific - Explosive Regional Growth

With Singapore constrained, demand is spilling into Malaysia, Indonesia, Thailand, and

India - one of the most dynamic shifts in the global market.

Israel - Can We Join the Race?

Yes - but only if we move decisively.

Israel’s advantages:

• Strategic location along major fiber routes
• Rapidly growing domestic AI demand
• Strong engineering capabilities
• Global cloud providers seeking a deeper regional footprint
• Highly skilled technical workforce

Conclusion

Hyperscalers - from AWS, Google, Microsoft and Meta to fast-growing AI clouds -

have become the central force shaping global digital-infrastructure development.

According to DC Byte, they generate nearly 70% of global data-center demand,

driving the acceleration of the AI revolution.

The biggest challenge is power. AI-focused facilities require tens or hundreds of

megawatts, and grids in markets like Northern Virginia, Frankfurt, Singapore, and

Amsterdam are already overextended.

The second force is regulation. Some countries are tightening environmental

requirements, while others are opening the door to rapid expansion- creating a

migration of hyperscale projects toward more supportive markets.

The third force is speed. Hyperscalers must secure capacity immediately, which is

why most new global capacity is pre-leased before a site even breaks ground.

Israel has a genuine window of opportunity: Despite grid limitations, its geography,

connectivity, and growing AI ecosystem position it to become a regional AIinfrastructure

hub. But this requires national strategy, available power, and rapid

infrastructure planning.

For CTOs, CIOs, and IT leaders:

Cloud performance and resilience are now directly tied to global infrastructure

availability. Multi-cloud and multi-region strategies have become essential, not

optional.

Key challenges ahead:

Power availability, Planning and regulatory bottlenecks, Shortage of technical

workforce and Lack of a modern cloud-infrastructure policy.

If Israel develops a national data-center plan — with strategic zones and long-term

electrical planning — it can evolve not only into a consumer of AI, but into a regional

production hub for AI infrastructure.