WuBlockchain’s WhiteLine Daily highlighted four developments across the AI sector, covering cloud infrastructure growth, software monetization, government-backed financing for compute supply chains, and a new inference-chip partnership tied to Nvidia’s platform strategy.
Oracle adds more than $30 billion in AI cloud contracts
Oracle, the U.S. enterprise software and cloud computing company, reported its latest quarterly results with cloud infrastructure revenue up 121% year over year to $7.4 billion. AI training and inference demand continued to drive expansion in that business.
According to Reuters, Oracle added more than $30 billion in AI cloud contracts during the quarter. The report said most of those contracts do not require substantial additional capital investment, which allowed the company to keep its annual capital expenditure target unchanged.
Free cash flow for the quarter was negative $5.4 billion, better than the market expectation of negative $9.56 billion, though it still reflected the funding pressure tied to data center expansion.
WhiteLine Daily’s takeaway was that if most of the newly signed AI contracts can be handled without major incremental capital spending, Oracle’s existing infrastructure buildout is beginning to absorb more demand, easing part of the tension between fast revenue growth and heavy cash consumption.
Adobe’s AI-first ARR grows more than 150%
Adobe also released its latest quarterly earnings. Revenue came in at $6.76 billion, above the market expectation of $6.7 billion.
Barron’s reported that annual recurring revenue, or ARR, from Adobe’s AI-first products rose more than 150% from a year earlier, showing continued expansion in paid revenue tied to those products. The report also made clear that this ARR growth rate applies to AI-first products only, not to Adobe’s total company revenue growth.
For the next quarter, Adobe guided revenue to a range of $6.8 billion to $6.85 billion. The midpoint was slightly below the market expectation of $6.85 billion, and the stock fell about 2% in after-hours trading.
WhiteLine Daily said Adobe’s AI product revenue is growing quickly, but that pace has not yet pushed the company’s overall revenue guidance above market expectations. Whether AI commercialization can translate into a more visible improvement in total company growth remains a key valuation test.
Report says Pentagon is discussing a roughly $5 billion loan for Fluidstack
The Wall Street Journal, citing people familiar with the matter, reported that the U.S. Department of Defense is in talks through its Office of Strategic Capital to provide roughly $5 billion in loans to AI cloud compute company Fluidstack.
Fluidstack mainly provides the computing infrastructure needed for AI training and inference. The proposed funding would be used to strengthen supply chains and manufacturing capacity for data center-related components in the United States, rather than to directly build a new AI data center.
If completed, the deal would be the largest loan ever made by that office. The transaction has not been finalized, and neither the Department of Defense nor Fluidstack had responded to the report at the time it was cited.
WhiteLine Daily’s view was that, if the loan goes through, U.S. government funding would extend support for AI infrastructure beyond compute deployment and into the equipment manufacturing side of the data center supply chain.
d-Matrix ties its next chip to Nvidia’s interconnect and rack system
AI inference chip company d-Matrix announced a partnership with Nvidia under which its next-generation Raptor chip will connect to the NVLink Fusion interconnect platform and the MGX rack system. The related products are expected to make their first appearance in the fourth quarter of 2027.
Under the company’s published design, Nvidia GPUs would handle compute-intensive tasks such as processing input content, while Raptor would take on the step-by-step generation work that requires faster response times. The target use cases include AI coding assistants, real-time chat, and voice agents.
Raptor is expected to tape out before the end of this year and has not yet entered formal delivery.
WhiteLine Daily said the arrangement creates a division of labor between a dedicated inference chip and Nvidia GPUs. Customers gain another chip option while still staying on Nvidia’s interconnect and rack platform, which could broaden the reach of that system ecosystem.
Main thread of the day
WhiteLine Daily said the market is now testing AI investment in more concrete ways: whether compute suppliers can use existing infrastructure spending to take on more orders, and whether software companies can turn AI product growth into an improvement in overall financial performance. Beyond revenue alone, equipment supply and inference efficiency are also shaping the pace of expansion, with financing structures and chip-level task allocation adjusting around those constraints.