Recently, new details about the StarGate Initiative have surfaced. This ambitious project was initiated by OpenAI with the goal of establishing an artificial intelligence data center cluster valued at up to 500 billion US dollars. However, a report from the Financial Times on Thursday indicated that the project currently lacks a fully mature plan and funding has yet to be finalized.
The StarGate Initiative was jointly unveiled by OpenAI, SoftBank Group Corp., and Oracle Corp. at a press conference held in the White House. These companies aim to build an extensive network of data centers specifically optimized for running AI workloads. Notably, sources cited by the Financial Times revealed that OpenAI will be the sole user of this data center cluster.
Shortly after the announcement of the StarGate Initiative, Elon Musk, founder of Tesla, criticized it. Musk is currently suing OpenAI for alleged breach of contract and runs a competing AI startup called xAI Corp., which focuses on large language model development. In a post on xAI Corp.'s platform, Musk stated that one of the main backers of the StarGate Initiative, SoftBank, does not have sufficient funds to cover the budget for this massive project. He bluntly commented: "They actually don't have money; SoftBank's margin is far less than 10 billion US dollars." In response, Sam Altman, CEO of OpenAI, quickly replied: "You know it's not true."
According to a report by the Financial Times on Thursday, OpenAI and SoftBank will initially invest around 15 billion US dollars into the StarGate Initiative, which is only a small portion of their planned investment of up to 500 billion US dollars in new data centers. The companies plan to fund the remainder of the project through equity financing from "existing investors" and debt financing. Additionally, the report noted that the plan will not receive government funding support.
Bloomberg also uncovered new details about the data centers being built by OpenAI and its partners on Thursday. Citing regulatory filings, the publication reported that the StarGate Initiative is currently constructing ten facilities in Abilene, Texas. OpenAI informed the newspaper that the company's 875-acre data center campus in the city "will be our first nationwide construction".
This week, besides OpenAI's major data center investment announcement, Meta Platforms Inc., developer of the open-source LLM series, disclosed its capital expenditure plan. Meta stated that its capital spending this year would reach between 60 billion and 65 billion US dollars, with most of the funds going towards building data centers and hiring AI experts. Mark Zuckerberg, CEO of Meta, revealed this information in a Facebook post, noting that the plan would add one billion watts of computing power within the next 12 months, enough to power approximately 700,000 households. By the end of this year, Meta expects to run over 1.3 million graphics processing units in its data centers.
Zuckerberg also outlined Meta's AI software roadmap. He mentioned that later this year, Meta plans to launch a new product in the open-source LLM series called Llama 4. Internally, the company will introduce an AI-driven programming assistant expected to contribute increasing amounts of code to the company's software projects.
As the StarGate Initiative and Meta's data center investment plans progress, the development of the AI field is undoubtedly set to enter a new peak period. However, ensuring adequate funding and smooth implementation of these massive projects, as well as maintaining a leading position in the fiercely competitive market, are significant challenges facing these companies.
