Energy, not chips, will determine AI supremacy — and here, it's advantage Gulf

Indian Express

21-05-2025

For the past decade, policymakers and analysts have been fixated on the semiconductor race, viewing the export controls of chips as the linchpin of artificial intelligence supremacy. The Biden administration's sweeping export controls aimed at curbing China's chip development have only reinforced this belief. Yet, while chips are undeniably critical, this fixation misses the deeper reality shaping the future of AI: The fundamental constraint on AI is not chips but energy.
As compute power has become central to AI's exponential growth, energy requirements have surged exponentially as well. According to a recent RAND report, 'training could demand up to 1 GW in a single location by 2028 and 8 GW — equivalent to eight nuclear reactors — by 2030, if current training compute scaling trends persist.' It's not surprising that energy costs now constitute the largest recurring operational expense for hyperscale data centers and AI clusters. Chips matter — but without abundant, affordable, and reliable energy, even the most advanced semiconductors are severely constrained.
This shift in the AI paradigm positions the Gulf states — Saudi Arabia and the UAE, in particular — in a central role in the emerging global compute ecosystem. The region's abundant and relatively low-cost energy supplies, coupled with aggressive investment strategies in digital infrastructure, uniquely position it to become the global backend for compute-intensive AI tasks. This is not theoretical speculation; it is already underway.
The recent visits by American tech CEOs to the Gulf, alongside President Trump on his high-profile Gulf tour, further underscore this emerging dynamic. Deals worth billions — Nvidia supplying thousands of advanced AI chips to Saudi Arabia's newly established AI champion, Humain, and Amazon's $5 billion joint AI zone — are merely the tip of the iceberg. What is truly noteworthy is not just the chips changing hands, but the scale of compute infrastructure these investments are building in the region.
Additionally, the US and UAE have announced a groundbreaking plan to develop a 5GW AI data center campus in Abu Dhabi, built by Emirati AI firm G42 and operated in partnership with American hyperscalers. The first phase alone will span 1GW and leverage nuclear, solar, and gas power. This facility, covering 10 square miles, is designed to serve as a regional platform enabling US hyperscalers to deliver AI services to nearly half the global population. The project reflects the UAE's ambition to lead as a global hub for AI research and innovation, highlighting significant American investments — including Microsoft's $1.5 billion stake in G42 and strategic partnerships with other US tech firms.
The Gulf states have the 'compute triangle'— abundant and inexpensive energy, scalable and sovereign-owned data infrastructure, and ample investment capital. No other region offers such an integrated value proposition for the demanding AI industry. To build AI infrastructure at global scale, a country or region needs more than advanced processors: It needs cheap, stable extensive energy infrastructure, and a leaner regulatory environment — all areas where the Gulf excels.
Critically, the energy reality reshapes the global calculus around AI. While the US has maintained a strategic posture largely focused on controlling semiconductor supply chains and restricting Chinese access, it risks overlooking the long-term imperative: Securing energy-rich partners to sustain its AI ambitions. In this context, US–Gulf relations must move beyond oil-for-security toward compute, where access to Gulf-based energy and data infrastructure becomes central to American technological competitiveness.
The rise of Huawei's advanced AI chips has only accelerated this urgency. Rather than simply chasing a technological blockade — an approach that is increasingly unsustainable — the United States must engage more deeply with energy-abundant allies who possess the capacity and the willingness to co-develop compute infrastructure. In essence, the Gulf offers an opportunity to maintain US dominance in AI by providing the very foundation that advanced semiconductors require: Immense quantities of affordable energy.
Yet, many in Washington remain preoccupied with the semiconductor rivalry alone. The Biden administration's semiconductor export controls were certainly necessary and have significantly impeded China's immediate technological ambitions. But these measures are temporary fixes. The fundamental drivers of technological leadership in AI will increasingly revolve around where compute infrastructure is sited, who controls energy inputs, and how efficiently that energy can be utilised.
The lasting strategic advantage will belong to those who control not just chip fabrication but also the energy-intensive infrastructure that powers AI training and inference at scale.
The Gulf states are not waiting for Washington to come around to this realisation. With their sovereign wealth funds, immense energy resources, and rapidly growing digital infrastructure, they are moving decisively toward becoming a global backend for AI compute. The real question facing American policymakers is whether they will seize this moment to establish a comprehensive and forward-looking AI-energy partnership with Gulf allies — or continue assuming a rapidly diminishing leverage over global compute.
Energy, not chips, is the real bottleneck for AI, and the Gulf states hold the keys. Washington must act swiftly to deepen its engagement and secure a lasting strategic partnership built not on oil alone but on compute, the fuel of the digital century.
The writer is a senior fellow at the Middle East Institute, a member of McLarty Associates, and a non-resident senior fellow at the Foreign Policy Research Institute