Yazılar

SoftBank Profit More Than Doubles to $16.6 Billion on OpenAI Valuation Gains

/in /tarafından

SoftBank Group reported a stunning surge in quarterly profits, more than doubling its net income to 2.5 trillion yen ($16.6 billion) in the July–September period, thanks largely to massive valuation gains from its stake in OpenAI, the creator of ChatGPT.

The figure far exceeded analyst expectations — three LSEG analysts had forecast an average profit of just 207 billion yen — and also dwarfed the 1.18 trillion yen profit recorded during the same period last year.

SoftBank’s Vision Fund unit, which manages the company’s global technology investments, posted a 3.5 trillion yen investment gain, with 2.16 trillion yen attributed directly to its OpenAI holdings.

The result comes amid a surge in AI-related stocks and infrastructure spending, pushing SoftBank’s shares to record highs. The company has emerged as one of the biggest beneficiaries of the AI investment boom, fueled by global demand for computing power and data centers.

In March, SoftBank led a $40 billion funding round valuing OpenAI at $300 billion. By October, it joined a group of investors purchasing $6.6 billion worth of OpenAI shares from employees at a $500 billion valuation, marking one of the largest private valuations in tech history.

Still, some investors are wary of an emerging “AI bubble”, questioning whether such vast capital inflows can sustain their expected returns.

SoftBank is also ramping up other AI and semiconductor bets. It recently sold 32.1 million shares of Nvidia for $5.83 billion, raised more than 620 billion yen in bonds across three currencies, and secured bridge loans totaling over $15 billion to fund its OpenAI and Ampere chip ventures.

Founder and CEO Masayoshi Son, known for high-stakes investments in transformative technologies, remains confident in AI’s potential despite a mixed record that includes triumphs like Alibaba and failures such as WeWork.

Amazon’s $38 Billion OpenAI Deal Signals Major Comeback in the AI Race

/in /tarafından

Amazon has struck a $38 billion cloud deal with OpenAI, marking a significant win for the company’s Amazon Web Services (AWS) division and a major step toward reclaiming lost ground in the artificial intelligence boom. The agreement comes after Amazon had faced mounting criticism for lagging behind rivals Microsoft and Google in securing AI partnerships and deploying consumer-facing language models.

After years of dominance in the cloud industry, Amazon’s market share slipped to 29% by September — down from 34% before ChatGPT’s debut in 2022, according to Synergy Research Group. The new partnership with OpenAI, however, suggests AWS is regaining momentum. The deal will allow OpenAI to use Amazon’s infrastructure, including its custom-built Trainium chips, to train next-generation models.

Analysts said the collaboration, though smaller than OpenAI’s $250 billion commitment with Microsoft’s Azure or Oracle’s $300 billion deal, is strategically vital for Amazon. “It’s a key first step in Amazon’s effort to partner with a company that will spend over a trillion dollars on computing power in the coming years,” said Mamta Valechha of Quilter Cheviot.

The announcement sent Amazon’s shares up 5%, hitting a record high. The company has recently expanded its AI footprint, including the launch of Project Rainier, an $11 billion AI data center in Indiana where models from startups like Anthropic are being trained. CEO Andy Jassy is also pushing a leaner management structure to boost efficiency, as Amazon plans to spend around $125 billion in capital expenditures this year — outpacing Alphabet’s $93 billion.

Analysts expect the OpenAI partnership to increase AWS’s backlog by about 20% in the fourth quarter, potentially adding $40 billion in future revenue.

Applied Digital Strikes $5 Billion AI Infrastructure Deal with U.S. Hyperscaler

/in /tarafından

Applied Digital (APLD.O) announced on Wednesday that it has signed a $5 billion, 15-year lease agreement with a U.S.-based hyperscaler for 200 megawatts (MW) of capacity at its Polaris Forge 2 data center campus in North Dakota, solidifying its position as a major player in AI infrastructure development. The deal sent Applied Digital’s shares up 4% in premarket trading.

The agreement is expected to generate about $5 billion in contracted revenue over its term and reflects the surging demand for high-performance compute capacity driven by the rapid adoption of artificial intelligence applications. Tech giants and AI developers are racing to secure energy-intensive infrastructure capable of training and deploying advanced language and vision models.

With this latest contract, Applied Digital’s total leased capacity across its Polaris Forge 1 and 2 campuses now reaches 600 MW, marking a significant milestone in its expansion strategy. The company also recently finalized a separate 150 MW lease with CoreWeave (CRWV.O) earlier this year, underscoring its growing role as a key infrastructure provider for the AI ecosystem.

Applied Digital’s stock has soared more than 325% in 2025, buoyed by investor enthusiasm for companies building AI-ready data centers capable of handling the computational load required by large language models and generative AI systems.

Industry analysts say the deal highlights how AI infrastructure has become the new frontier of big tech investment, with hyperscalers — massive cloud computing companies such as Google, Amazon, and Microsoft — locking in long-term capacity agreements to meet explosive AI demand.

The company’s Polaris Forge complex in North Dakota is one of several U.S. projects focused on delivering high-density compute environments optimized for AI workloads. Applied Digital said the partnership will also support future energy efficiency improvements and renewable power integration, aligning with broader sustainability goals across the data center industry.