AetherGrid Digital Infrastructure Business Plan › Section 6 › Development Strategy & Campus Roll-Out

AetherGrid develops 155 MW across four phases, sequenced to prove the
model in Johannesburg before extending to Cape Town, Durban and national
expansion. Each campus is a discrete, financeable project with its own
land, power, anchor tenants and SPV, allowing capital to be committed in
disciplined tranches against secured power and demand.

Figure 7

6.1 Phasing logic

The phasing is deliberately proof-led. Project Atlas in Johannesburg,
the deepest market, closest to financial-services demand and cloud
on-ramps, establishes the operating track record, grid relationships,
anchor-tenant references and financing precedent that de-risk everything
after it. Project Horizon (Cape Town) captures subsea-cable connectivity
and renewable resource; Project Oceans (Durban) adds a third gateway and
disaster-recovery geography; and Phase 4 scales the platform nationally
to 155 MW once the model is proven. Capacity reaches commercial
operation roughly two years behind capital commitment, producing the
revenue J-curve modelled in Section 8.

6.2 Technology and design infrastructure

The design is explicitly AI-ready: high-density racks, liquid-cooling
capability and the electrical capacity to support GPU clusters, so
AetherGrid can serve the fastest-growing demand segment from day one
rather than retrofitting. Closed-loop cooling addresses the
water-scarcity constraint that makes conventional evaporative cooling
problematic in South Africa.

6.4 Reference technical architecture

Each campus is built to a standardised, modular reference
architecture that allows capacity to be released in phased blocks
matched to demand and power availability, while maintaining Tier III/IV
resilience. Modularity is the key discipline: it aligns capital
deployment with lease-up, limits stranded capacity, and lets the
platform scale power in the 5–10 MW increments the grid constraint
imposes.

The high-density AI zones are the architectural differentiator.
Rather than retrofitting liquid cooling into air-cooled halls, costly
and disruptive, AetherGrid designs liquid-cooling-ready zones from
inception, so it can accept GPU clusters at 30–80 kW per rack as AI
demand arrives. Combined with the modular power blocks and
captive-renewable supply, this positions each campus to serve the
highest-value, fastest-growing workloads without stranded legacy
design.

6.3 Sustainability and renewable-energy strategy

Sustainability is both an operating necessity and a commercial
differentiator. The renewable strategy targets 80% renewable energy by
Year 10 through on-site solar, battery storage and renewable
power-purchase agreements, mirroring the captive-renewable model the
market leader has adopted to secure predictable, clean power and
insulate against load-shedding. This lowers and stabilises energy cost,
meets the carbon requirements of global cloud tenants, and directly
addresses the power constraint that gates the whole sector.

Figure 8

Beyond energy, the design targets progressive improvement in power
usage effectiveness (PUE) toward 1.25 through free-air and liquid
cooling, and closed-loop cooling to minimise water draw in a
water-scarce country. These commitments are increasingly prerequisites,
not enhancements, for winning hyperscale tenants and for accessing DFI
and climate-linked capital.

Confidential — this business plan is provided to prospective investors and lenders for evaluation purposes only and may not be reproduced or distributed without the written consent of AetherGrid Digital Infrastructure Holdings (Pty) Ltd.