Solar lets them build their own supply at the pace they need.
À medida que a inteligência artificial expande suas raízes no tecido da economia digital, as empresas que a sustentam enfrentam uma questão antiga sob nova forma: de onde virá a energia para alimentar o pensamento das máquinas? A Meta respondeu com um contrato de 650 megawatts de energia solar firmado com a AES no Texas e no Kansas, seu quarto acordo do tipo em 2025, consolidando uma carteira global que ultrapassa 12 gigawatts de capacidade renovável. A escolha não é apenas ambiental — é estratégica, guiada pela velocidade de implantação e pelo custo da eletricidade que determinam quem lidera a corrida pela infraestrutura da IA.
- A demanda por energia dos data centers de IA cresce mais rápido do que a infraestrutura tradicional consegue acompanhar, criando uma pressão urgente por fontes de eletricidade rápidas e baratas.
- A Meta fechou quatro contratos solares em apenas cinco meses de 2025, todos no Texas, sinalizando uma concentração deliberada de capital em estados com licenciamento ágil e redes elétricas eficientes.
- O Texas lidera as instalações solares nos EUA pelo segundo ano consecutivo, atraindo investimentos bilionários de gigantes da tecnologia graças ao seu mercado de energia desregulamentado e clima favorável.
- A AES, desenvolvedora dos projetos, destaca que fazendas solares podem gerar eletricidade antes mesmo de estarem totalmente concluídas — uma vantagem de prazo decisiva para empresas que precisam escalar operações de IA rapidamente.
- Com mais de 12 gigawatts de capacidade renovável global, a Meta posiciona energia limpa não como virtude corporativa, mas como infraestrutura competitiva essencial para o futuro dos seus negócios.
A Meta anunciou esta semana um novo contrato de energia solar de 650 megawatts com a empresa AES, divididos entre o Texas (400 MW) e o Kansas (250 MW), para abastecer seus data centers em expansão acelerada — infraestrutura central para as operações de inteligência artificial da companhia.
Este é o quarto acordo solar fechado pela Meta em 2025, e todos os quatro foram firmados no Texas. Os contratos anteriores somaram mais de 1,5 gigawatt no estado, e a concentração não é coincidência: o Texas liderou as instalações solares nos Estados Unidos por dois anos consecutivos, impulsionado por sol abundante, processos de licenciamento ágeis e uma rede elétrica funcional — condições que fazem diferença real para quem precisa construir capacidade computacional em ritmo acelerado.
O que torna a energia solar especialmente atraente nessa escala é a velocidade. Diferentemente de usinas convencionais, fazendas solares podem começar a gerar eletricidade antes mesmo de estarem totalmente prontas. O CEO da AES, Andrés Gluski, apontou que a combinação de geração rápida e baixo custo é o principal fator que atrai empresas de tecnologia de grande porte para esse modelo. Para a Meta, que precisa alimentar milhares de servidores rodando modelos de IA, velocidade e eficiência de custo se traduzem diretamente em vantagem competitiva.
Com mais de 12 gigawatts de capacidade renovável em todo o mundo, a Meta consolida uma aposta que vai além da responsabilidade ambiental: trata-se de garantir energia confiável e acessível no ritmo que o negócio exige. O padrão que emerge — quatro contratos em cinco meses — deixa claro que a empresa já fez sua escolha sobre como alimentar a próxima fase da inteligência artificial.
Meta announced a new solar energy contract this week, securing 650 megawatts of capacity across Texas and Kansas through the power company AES. The deal breaks down to 400 megawatts in Texas and 250 in Kansas, both aimed at feeding the company's rapidly expanding data centers—infrastructure that has become essential as Meta scales up its artificial intelligence operations.
This is Meta's fourth solar agreement signed in 2025, and all four have landed in Texas. The previous three contracts brought 595 megawatts, 505 megawatts, and two separate 200-megawatt projects to the state. The pattern is deliberate. Texas has led the nation in new solar installations for two consecutive years, 2023 and 2024, a position built on three structural advantages: abundant sunshine, streamlined permitting processes, and grid connections that work. For a company racing to build out computing capacity, these conditions matter enormously.
What makes solar particularly attractive to companies operating at Meta's scale is the speed of deployment. Unlike traditional power plants, solar farms can begin generating electricity before construction is fully complete—a timeline advantage that AES, the project developer, emphasized. Andrés Gluski, the company's CEO, noted that rapid energy generation combined with low electricity costs are the primary factors drawing hyperscale technology firms to solar contracts. For Meta, which needs to power thousands of servers running AI models, both speed and cost efficiency translate directly to competitive advantage.
Across all its projects globally, Meta now holds more than 12 gigawatts of renewable energy capacity. That figure represents a substantial commitment to clean power infrastructure, though it also reflects the sheer scale of computational demand that modern AI systems create. Each data center consumes enormous amounts of electricity; each new AI model trained or deployed requires more. The company's solar strategy is not primarily about environmental virtue—it is about securing reliable, affordable power at the pace and scale the business requires.
The Texas focus also signals something broader about how American infrastructure is reshaping itself around technology. States with favorable regulatory environments and existing grid capacity are attracting the largest capital investments from the companies that drive the digital economy. Texas, with its deregulated energy market and business-friendly permitting, has become the natural hub for this expansion. Other states are watching, and other companies are following similar patterns.
Meta's renewable energy deals are part of a larger industry trend. As AI workloads grow exponentially, the companies building the infrastructure to support them face a choice: secure power however they can, or invest in clean energy sources that offer both cost advantages and the speed of deployment they need. Meta's four solar deals in five months suggest the company has made its choice clear.
Notable Quotes
Rapid energy generation and low electricity costs are the primary factors drawing hyperscale technology firms to solar contracts— Andrés Gluski, CEO of AES
The Hearth Conversation Another angle on the story
Why does Meta need so much solar power all at once? Couldn't they just buy electricity from the grid like everyone else?
They could, but the grid can't deliver what they need fast enough or cheaply enough. AI data centers consume staggering amounts of power—we're talking hundreds of megawatts per facility. If Meta tried to source that through traditional utilities, they'd face years of waiting, higher costs, and no guarantee of capacity. Solar lets them build their own supply.
But solar is intermittent. The sun doesn't shine at night. How does that work for a data center that runs 24/7?
That's the practical reality, yes. What Meta is doing is layering sources—solar for the daytime load, grid power and other renewables for night hours. The 650 megawatts they're securing isn't their total power need; it's a major piece of it. And solar is still cheaper than the alternative.
Why Texas specifically? There are sunny places all over the country.
Texas has three things other states don't: a deregulated energy market that moves fast, permitting processes that don't take five years, and grid infrastructure already built to handle large industrial loads. A solar farm in Texas can go from approval to generating power in months. Anywhere else, you're looking at years of bureaucracy.
Is this good for Texas, or is Meta just extracting value?
Both, probably. Texas gets tax revenue, construction jobs, and a major tech employer. But Meta gets what it really wants—cheap, fast power to run AI systems that generate enormous profits. The state benefits, but the company benefits more.
What does this mean for other tech companies?
It's a signal. If you're building AI infrastructure and you're not securing renewable power contracts, you're falling behind on cost and speed. Every other hyperscaler is watching what Meta does. Within a few years, this won't be unusual—it'll be standard practice.