Unbiased AI-powered news
Dallas-based DataBank has obtained $2 billion in construction financing for three data center buildings in Red Oak, Texas, with plans for a fourth building funded by an additional $600 million. The campus, leased to a major hyperscaler, targets growing demand for AI inference computing. Deliveries are expected from 2026 to 2027.
Substrate placeholder — needs reviewDallas-based data center developer and operator DataBank has secured $2 billion in construction financing from a group of banks led by Mitsubishi UFJ Financial Group to build three data center buildings in Red Oak, a suburb 20 miles outside Dallas. The company did not disclose the other banks involved in the loan.
Business Insider reported that this financing supports a project amid surging demand for AI computing facilities closer to population centers. MUFG is leading a separate effort to raise roughly $600 million in debt for a fourth building at the Red Oak campus, sourced from the private placement market tapping Wall Street investors.
The four-building cluster has been leased by a hyperscaler, identified by DataBank CEO Raul Martynek as one of Amazon, Google, Microsoft, Oracle, or Meta. Martynek declined to name the specific tenant.
The four buildings will total 240 megawatts of computing capacity and comprise half of an planned 8-building, 480-megawatt data center campus in Red Oak. DataBank expects to deliver the first of the four initial buildings in the third quarter of 2026.
The last of these buildings is slated for completion by the end of 2027, according to Martynek. Texas has emerged as a hot spot for large-scale data centers in remote locations that handle artificial intelligence training. DataBank's Red Oak project reflects a shift toward facilities nearer to urban areas for faster AI model delivery to users.
"Why do data center users want to be close to metro markets? Because that's where most of the fiber is and that's where people are," Martynek said. "Who consumes these products?
In 2025, 9% of data center workloads were for inference computing, while 14% were for training and 77% for traditional cloud computing, according to data from JLL. JLL predicts that by 2030, 37% of the industry's workloads will be for inference computing and 13% for training.
Carl Beardsley, head of data centers for JLL Capital Markets, stated, "It's on the forefront of a lot of minds. " Martynek noted challenges in securing financing due to a slowdown in the syndication market, with lenders cautious about heavy exposure to AI and data center sectors.
He explained that the $600 million for the fourth building was separated into a private placement to navigate these conditions. "It took us longer than we originally anticipated because obviously we're part of the market," Martynek said. "If a major money center bank says, 'Hey, you know what?
TankerTrackers data shows 36 million barrels shipped and another 36 million still at sea. Iranian officials separately reported 25 million barrels crossing the blockade line since Monday.
ForbesUFC CEO Dana White stated that negotiations for a cage fight between Elon Musk and Mark Zuckerberg were genuine and included discussions about holding the event at Rome's Colosseum. White said the venue requested an estimated $150 million, which would have gone toward restoring o…
foxnews.comGround beef prices have risen more than 20 percent since January 2025. U.S. and Mexican negotiators met June 16-17 to discuss the trade deal while President Trump warned Washington may withdraw.