Вот английская версия — ясная, аккуратно собранная и без лишних украшений.
**OpenAI’s Energy Ambitions: A Data-Center Plan Comparable to India’s Power Use**
OpenAI CEO Sam Altman has outlined an extraordinarily ambitious target: by 2033, the company aims to expand its computing capacity to **250 GW**. This figure has been widely compared to the electricity consumption of India — not India’s *current* total use, but a comparable scale of installed or required capacity.
Reports indicate that such infrastructure would require up to **60 million Nvidia GPUs**, with **30 million new GPUs each year** just to keep the system running. The projected cost reaches an astonishing **$12.5 trillion**.
Environmental impact assessments suggest that CO₂ emissions from this future network of data centers could be roughly **twice as high** as those of ExxonMobil — one of the largest private carbon emitters in the world.
Altman describes this trajectory as a form of “brutal industrialization,” portraying it as the necessary road toward **AGI (Artificial General Intelligence)**.
**Bibliography**
Tom’s Hardware – “OpenAI’s colossal AI data center targets…”
iXBT – Coverage of OpenAI’s 2033 250 GW infrastructure plans
BTCC – Cost estimates reaching $12.5 trillion
Fortune – Interview: Altman on the “gigawatt arms race”
**Hashtags**
#OpenAI #SamAltman #AIInfrastructure #DataCenters #GPUEconomy #Nvidia #AI2033 #ArtificialGeneralIntelligence #EnergyCrisis #ClimateImpact #TechIndustry #AIPowerDemand #IndustrialScaleAI #CarbonEmissions
Yonhap Infomax News
Pure Acetone