Citi sharply raised its capital-spending forecasts for Alibaba, reiterating a Buy rating and $190 price target as the company targets 20 gigawatts of AI capacity by 2032. Citi analyst Alicia Yap wrote that assuming a 50/50 Capex/Opex partner split at US$30bn per 1GW, the firm lifted its FY27 through FY29 estimated capex to Rmb258bn, Rmb283bn and Rmb282bn. Under those assumptions, external AI cloud revenue could reach $168 billion by fiscal 2033, roughly 40% compound annual growth from fiscal 2026, with a sensitivity range of $126 billion to $210 billion. Alibaba is also broadening its technology push: T-Head unveiled its Zhenwu V900 accelerator and Lingjun Supernode infrastructure to power AliCloud's production-ready agentic cloud platform, and the company is expanding its Qwen model family, including the 2.4-trillion-parameter Qwen3.8-Max, while developing smaller models for cheaper deployment and moving into image, audio and world models. Alibaba shares nevertheless fell 3.7% in premarket trading amid concerns surrounding a reported Chinese government probe.
CoreWeave Adds Harell Data as Multiyear AI Infrastructure Customer
CoreWeave has signed a multiyear agreement with Harell Data, which will use the company's cloud platform to run model training, fine-tuning and inference workloads. Under the deal, the workloads will run on Nvidia A100 and Hopper GPUs through CoreWeave, while Harell said the raw data itself will remain inside its own platform. Harell Data provides proprietary scientific datasets used to train AI models, and its platform lets customers run models against those datasets without receiving direct access to the underlying data. The agreement reinforces CoreWeave's position as a specialized provider of high-performance AI compute and gives investors another data point supporting its push to lift revenue generated from each gigawatt of deployed infrastructure. UBS recently initiated coverage of CoreWeave with a Buy rating and a $120 price target, arguing that revenue per gigawatt could rise from roughly $11 billion currently to more than $15 billion over time. CRWV shares were up about 1% in premarket trading Wednesday, and the stock has gained roughly 20% year to date.
Nscale Files for US IPO, Backed by $1 Billion From Nvidia
Nscale filed for a U.S. initial public offering on September 18, disclosing a $1.02 billion loss in the first six months of 2026 on just $140.6 million of revenue. NVIDIA Corporation is committing $1 billion to a $3.1 billion convertible financing, and Microsoft Corporation is expected to become one of Nscale's most important customers. As of August 31, Nscale had $103.4 billion of active and contracted total contract value under long-term take-or-pay contracts, with a weighted-average contract life of about 5.7 years, operating in 14 regions with more than 10 gigawatts of potential power capacity. One customer generated 52% of first-half revenue, and the company's future plan leans heavily on Microsoft and Anthropic. Insider Monkey's database counted 285 hedge funds with reportable Nvidia longs at the end of Q2 2026, up from 275 in Q1, while Microsoft holders fell to 273 from 282.
Microsoft to Spend $10 Billion in Gulf Through 2030, Smith Says
Microsoft will spend $10 billion through 2030 across four Persian Gulf countries, building out infrastructure and strengthening cyber security, Microsoft President Brad Smith told Bloomberg's Ed Ludlow. Smith said the commitment reflects confidence in the Gulf nations, which he said are leading the world in AI transformation, and includes a digital resilience initiative to strengthen cyber security amid conflict in the region. He confirmed Microsoft remains committed to its earlier pledge to use 200 megawatts of capacity in Abu Dhabi, and said a new set of cloud services will go live in Saudi Arabia in just five weeks. On AI safety, Smith said safety will not advance if the world relies on only one or two companies or a single slice of the AI sector, arguing that safety should be built in at multiple layers, including model companies, software firms like Microsoft, and hyperscalers, and that independent evaluators must be capable, properly chosen, and genuinely independent. He noted Microsoft sits on a joint deployment safety board with OpenAI that must weigh in before the next major model is released, and said Microsoft updated its ethical guidelines last week for what he called the age of AI.
Alibaba Group Holding Ltd
Astera Labs, Inc.
Citigroup Inc.