ComputeLabs Research
Edition of · 19 stories
The supplied reports identify Chief Financial Officer Colette Kress as NVIDIA’s interim chief executive following Jensen Huang’s resignation. A Financial_Express summary also reports that Huang resigned as president and director, while the company initiated a search covering internal and external candidates for the permanent CEO position.
Corvex’s SEC-filed November 12 press release discloses an actual $7.5 million borrowing, rather than merely an available financing commitment. The subsidiary financing with USD.AI funds GPU servers for the Blackwell expansion at Mid-Atlantic 1; the supplied excerpt does not specify the facility’s maximum size or the number of servers being financed.
The disclosed collateral package covers the borrowing subsidiary’s assets, expressly including the related customer contract. The financing therefore includes contractual assets as collateral, rather than being described solely as a pledge of GPU hardware.
RUM Group’s quarterly filing dates the initial Northern Data acquisition closing to June 17, 2026, when it acquired approximately 85% of outstanding shares. On September 30, it acquired additional shares from Tether under the Transaction Support Agreement, increasing ownership to approximately 98%.
A financialjuice message attributing its report to CNBC states that OpenAI raised $15 billion at a $1.25 trillion valuation. A separate Financial_Express market summary states that the initial financing target was $30 billion.
A Financial_Express message attributes the June 15 tape-out milestone to Cerebras’ chief executive. The statement concerns the next-generation Wafer Scale Engine 5 accelerator; it does not announce customer shipments or commercial availability.
Applied Materials reported fiscal Q4 net sales of $10.41 billion, against the $10.27 billion analyst estimate cited by financialjuice. Semiconductor Systems contributed $8.00 billion, compared with the cited $7.94 billion estimate.
Cisco’s revised $9 billion target covers fiscal 2027 AI-infrastructure revenue from hyperscalers, not total company revenue. Fiscal Q1 revenue was $18.50 billion, compared with the $18.13 billion estimate cited by financialjuice.
Financial_Express reports that SpaceX announced the Arizona investment plan on its website on November 12. The $50 billion project is located in Cochise County and is distinct from the proposed computing facility near the company’s Starbase headquarters in Texas.
Cerebras’ CEO said the company had secured approximately 700 megawatts of total data-center capacity, with delivery scheduled by the end of 2027. The figure describes data-center capacity, not accelerator count, computing throughput, or manufacturing output.
Fermi Inc.’s quarterly filing states that it had not commenced revenue-generating activities as of September 30, 2026. Activity through that date concerned formation, infrastructure procurement, leasing discussions, preliminary site development, and marketing for Project Matador, its Texas campus.
A financialjuice message states that, as of September 30, Nscale Limited expected approximately $19 billion–$21 billion in total capital spending to finish development of its contracted data-center sites. The scope is the contracted-site development program, not a single campus.
The supplied reports describe disruption to Microsoft 365, Outlook, Teams, and Copilot during much of the November 12 U.S. working day. Microsoft said its investigation centered on Exchange Online, with some servers unexpectedly becoming unavailable and increasing load on the remaining infrastructure.
A Financial_Express report quotes Cerebras CEO Andrew Feldman saying Microsoft signed the agreement in October. Management expected revenue to begin in Q1 2027, distinguishing the signed partnership from its anticipated revenue commencement.
Cerebras’ CEO attributed the delayed initial OpenAI compute-capacity deliveries to delays in data-center capacity contracts. The supplied statement identifies a contracting-related delivery issue; it does not attribute the delay to accelerator design or manufacturing.
Bit Digital’s quarterly filing discloses that WhiteFiber entered into a master service agreement and initial service order with a technology customer in September 2026. The deployment uses approximately 1,200 NVIDIA B300 GPUs at a third-party data center in Sydney, Australia; this is a GPU count, not a server or rack count.
A financialjuice message reports that Google brought Gemini 4 Hydrogen to Search globally. The specified product change is the addition of visual and table-based AI responses within Search.
Anthropic’s announcement, relayed by Financial_Express, introduces Claude Meetings as a beta feature for Team and Enterprise plans. Claude can join a video call as a visible participant.
A press release filed by HeartCore Enterprises (HTCR) describes ONE AI’s Axis.ONE platform as supporting inference on Intel central processing units (CPUs), neural processing units (NPUs), and other edge hardware without dedicated GPUs. It is designed for 100% on-premise operation, including air-gapped and closed-network environments without cloud dependency.
