Cheapest Solar on Earth Can't Power AI in 50°C Heat: Here's Why
If you are a procurement officer managing infrastructure spend in the Gulf, an operator evaluating data centre investments in Saudi Arabia or the UAE, or a trade professional watching the China-MENA energy corridor take shape, the gap between the region’s solar potential and its AI ambitions is the most important number you are not tracking. This article explains why cheap energy is not the same as usable energy, and what China’s dominance in solar, battery, and grid technology means for your next project.
Table of Contents
- What China Built While Nobody Was Watching
- The Gulf Has the Cheapest Energy on Earth. That Is Not the Same as Having a Grid.
- The Problem Is Not Generation. It Is Everything That Comes After.
- The Rest of MENA Is Not Even in This Conversation
- Where China Sits in All of This
- What This Means for Anyone Operating in the Corridor
- FAQ
- Related Reading
What China Built While Nobody Was Watching
The numbers are not debatable at this point.
Since 2021, China has added more power capacity across all energy technologies than the United States has built in its entire history. In 2025 alone, China added 543 gigawatts. It plans to add 3.4 terawatts over the next five years, nearly six times as much as the US under BloombergNEF’s base-case scenario. By March 2026, China’s total installed generation capacity stood at 3.96 billion kilowatts. Non-fossil energy now accounts for 62% of that. Solar capacity grew 31.3% year on year to 1.24 billion kilowatts. Wind grew 22.4% to 660 million kilowatts.
China’s data centre sector consumed 22.9 billion kilowatt-hours of electricity in Q1 2026, up 44% year on year. China is adding 8 gigawatts of new data centre capacity this year alone, on track to reach 60 GW by 2030.
China did not build this to win an AI race. Xi Jinping did not wake up in 2022 and say “let’s do renewables for ChatGPT.” The renewable buildout started in the 1980s, Western firms trading market access for technology transfer, one five-year plan after another. The AI dividend is an accident of consistent long-term industrial policy. China built the grid. Then AI arrived and needed exactly what China had already built.
The United States, as former Treasury Secretary Hank Paulson and former Ambassador Nick Burns explain, took its eye off that ball. The Middle East never had the ball to begin with. That is both the problem and the opportunity.
The Gulf Has the Cheapest Energy on Earth. That Is Not the Same as Having a Grid.
Here is what surprises most people who look at Middle East energy data for the first time.
Gulf electricity tariffs run between $0.05 and $0.06 per kilowatt-hour. The US average is $0.09 to $0.15. Europe is higher still. The Gulf is not expensive to power. It is among the cheapest places on earth to generate electricity, sitting on the world’s largest solar irradiance resource, with flat land, no shading, no permits that take seven years, and sovereign wealth to fund construction without debt markets.
MENA renewable capacity jumped 44% in 2025, reaching 43.7 GW. Solar PV led with 34.5 GW. The region’s project pipeline has reached 202 GW, nearly matching aggregated national targets out to 2030. In 2025, MENA set global records. Solar PV tender prices dropped to 1.09 US cents per kilowatt-hour. Wind hit 1.33 cents. These are the cheapest clean energy prices ever recorded anywhere on earth.
So the Gulf has cheap energy, abundant sun, massive capital, and political will. It should be the easiest place on earth to power AI infrastructure.
It is not. And the reason is specific.
The Problem Is Not Generation. It Is Everything That Comes After.
Transmission and distribution losses in parts of MENA remain almost double global averages. Hyperscale AI facilities add concentrated, high-density demand to grid systems designed for predictable, distributed loads. Cooling already accounts for up to 70% of residential electricity consumption in some Gulf markets.
Then there is the heat.
Traditional data centre cooling systems rely on water temperatures around 4 to 6 degrees Celsius. When ambient temperatures exceed 50 degrees Celsius, standard summer conditions across the Gulf, air cooling fails entirely. Above 50 kilowatts per rack, which AI infrastructure routinely exceeds, air cooling becomes inefficient and liquid cooling becomes mandatory. In extreme Middle Eastern climates, cooling infrastructure can consume 50% or more of a facility’s total power.
Run the numbers. A Gulf data centre uses $0.05/kWh electricity. But half the electricity goes to keeping the servers from melting. The effective compute cost is now $0.10/kWh, comparable to the US. The cheap energy advantage disappears into a cooling system fighting 50-degree heat.
This is not a capital problem. Saudi Arabia has the money to build anything. It is an engineering problem specific to this geography that requires specific solutions: liquid cooling technology, water-free cooling systems, and subterranean siting, all approaches the region is only beginning to work through at scale.
The WEF notes that AI training, inference, and cloud services each carry distinct load profiles that existing Gulf permitting and grid forecasting models were not designed to accommodate. Properly integrated, hyperscale infrastructure can support renewable integration. Poorly integrated, it compounds grid stress.
The Gulf is currently in the “poorly integrated” phase. It is building fast. The grid integration work is lagging the construction pace.
The Rest of MENA Is Not Even in This Conversation
Everything above applies to the Gulf: Saudi Arabia, UAE, Qatar, Kuwait. Countries with capital, functioning grids, and sovereign ambition.
The rest of MENA is a different conversation entirely.
Lebanon runs on private diesel generators for 12 to 20 hours per day. The national grid delivers power for 2 to 4 hours. Syria is rebuilding its entire energy infrastructure from near-zero after 13 years of war. Iraq’s grid loses approximately 40% of generated power to transmission losses. Egypt is managing rolling blackouts in summer while simultaneously trying to build a renewable energy export corridor.
When the video asks “what does the US understand about AI and energy that it is getting wrong,” that is a question about policy, about prioritisation, about political will in a country that already has infrastructure. The equivalent MENA question is starker: large parts of this region do not have functioning electricity at all, let alone the grid stability that AI infrastructure requires.
The two-speed Middle East that Silk Road Intel has been mapping all year, Gulf AI ambition versus Levant and North Africa infrastructure deficit, runs on energy as its central axis.
Where China Sits in All of This
Here is the connection that makes this commercially important rather than just analytically interesting.
China accounts for roughly 80% of global solar technology production and battery technology, and more than 70% of wind technology. The Gulf’s 202-gigawatt renewable pipeline is built almost entirely on Chinese panels, Chinese inverters, Chinese battery storage systems, and Chinese construction contractors.
China just AI-mapped its entire renewable energy grid, giving grid operators what one researcher called a “God’s-eye view” of the national energy landscape. The optimisation tools China is now deploying to manage its own grid instability from variable renewables are the same tools the Gulf needs to manage its own integration problem.
The Gulf is not competing with China on energy infrastructure. It is buying from China to build its energy infrastructure. Every gigawatt of Gulf solar requires Chinese panels. Every gigawatt of battery storage requires Chinese cells. The rare earth truce discussed after the Beijing summit is not an abstract geopolitical event for Gulf project managers. It is a supply chain dependency that runs through every energy project on their books.
China plans to export its energy transition leadership. EV exports surged 80% in 2025, battery exports rose 40%, solar panel exports grew 20%. The Gulf is the buyer of significant portions of all three categories.
There is a phrase that captures what is actually happening here. China built the infrastructure. The US built the models. The Gulf is paying for both and trying to integrate them in 50-degree heat.
What This Means for Anyone Operating in the Corridor
The video Paulson appeared in frames the energy-AI race as a US-China competition. From Melbourne, looking at the China-MENA corridor, it looks like something different.
The Gulf has capital, cheap energy fundamentals, and procurement scale. China has the technology, the manufacturing capacity, and the grid management systems the Gulf needs. The bottleneck is not money or sunlight or political will. It is the engineering gap between cheap generation and reliable, AI-grade power delivery in an extreme climate, and the supply chain intelligence to build it at the pace Vision 2030 demands.
That gap is exactly where trade intelligence operates. Who makes the right cooling systems for desert AI data centres at Gulf procurement scale? Which Chinese manufacturers have desert-hardened battery storage at the spec Gulf operators need? What does the rare earth supply chain disruption mean for the 202-gigawatt pipeline sitting on Chinese panels?
These are not rhetorical questions. They have specific answers. Finding them is the work.
FAQ
Why can’t the Gulf just use its cheap solar to power AI data centres?
Cheap generation is only half the equation. Gulf data centres lose up to 50% of total power to cooling infrastructure fighting 50-degree ambient heat. The effective compute cost doubles from $0.05 to $0.10 per kilowatt-hour, erasing the energy advantage. The bottleneck is not the cost of electricity. It is the engineering of delivering reliable, AI-grade power in extreme climates.
What is the rare earth connection to Gulf energy projects?
China controls over 80% of global rare earth processing and more than 70% of wind and solar technology manufacturing. Every gigawatt of Gulf solar capacity depends on Chinese panels, inverters, and battery cells. The rare earth supply chain truce discussed at the Beijing summit directly affects the cost and availability of materials Gulf project managers need for renewable energy builds.
How does this affect procurement operators specifically?
Procurement officers sourcing cooling systems, battery storage, and solar components for Gulf data centres are buying almost exclusively from Chinese manufacturers. Supply chain disruptions, export controls, or price shifts in China directly affect project timelines and landed costs. Understanding which Chinese manufacturers have desert-rated spec at scale is a procurement advantage.
Is the rest of MENA (outside the Gulf) relevant to this story?
The Gulf states (Saudi Arabia, UAE, Qatar, Kuwait) have capital and functioning grids. The rest of MENA, Lebanon, Syria, Iraq, Egypt, faces a more fundamental problem: basic grid reliability. AI infrastructure requires stable, high-density power that most of the region cannot yet deliver. The two-speed energy landscape maps directly onto procurement risk.
What should trade operators watch for next?
Three signals matter: Chinese export policy on solar and battery technology, Gulf grid integration progress (whether permitting and forecasting models catch up to construction pace), and rare earth supply chain developments after the Beijing summit truce expires. Each one affects project costs and timelines for anyone building in the corridor. (China’s battery surplus meets Gulf storage demand)## Related Reading
- West Power Crisis: Middle East AI: How Gulf states are navigating the collision between AI infrastructure ambitions and grid capacity limits.
- Beijing Just Reshuffled the Global Trade Deck: The rare earth supply chain truce and what it means for Gulf project managers who depend on Chinese materials.
- The Battery Century Started in China: How China built dominant battery manufacturing capacity and what it means for energy storage procurement across MENA.
- China’s Different AI Race: Why China’s AI strategy is built on industrial infrastructure, not consumer apps, and why Gulf operators should pay attention.
Leo Houssami is the founder of Silk Road Intel, a trade intelligence platform covering the China-MENA-Australia corridor. Arabic and French fluent. Melbourne-based. silkroadleo.com
Credit: The framing of this article was inspired by a PBS NewsHour video featuring Hank Paulson and Nick Burns discussing what China understands about AI and energy that the US does not. The MENA substitution and trade corridor analysis are original.
Sources: Enerdata, China Renewable Capacity January-February 2026 (April 2026). Rystad Energy, China Data Center Capacity to Top 60 GW by 2030 (May 2026). Bloomberg/Energy Connects, China Ramps Up Energy Boom (February 2026). Dii Desert Energy MENA Energy Outlook 2026 (January 2026). WEF/SRMG Think, MENA Data Centres and AI Infrastructure (May 2026). Intelligent Data Centres, Middle East AI Ambitions and High-Power Battery Solutions (March 2026). GIS Reports, Middle East Energy Security (February 2026). Brookings, Global Energy Demands Within AI Regulatory Landscape (April 2026). China Daily, Solar Capacity to Outshine Coal 2026 (April 2026).
#SilkRoadIntel #MENAenergy #ChinaEnergy #AIinfrastructure #Gulf #Vision2030 #SolarEnergy #SupplyChain