Coverage window: November 12–15, 2026. Only material developments with sources published in this period are included.
Compute and infrastructure: Nvidia faces a leadership transition
Nvidia says Jensen Huang has resigned as CEO, president and board member, effective immediately. CFO Colette Kress is interim CEO while the board searches for a permanent successor. Reuters reports that the departure has raised questions about execution and the wider AI ecosystem Huang helped finance. Economically, this is a governance and continuity risk for a company central to accelerator supply and AI infrastructure investment—not evidence that its product roadmap or customer commitments have changed. ([nvidianews.nvidia.com](https://nvidianews.nvidia.com/news/nvidia-announces-ceo-transition?ftag=YHF4eb9d17))
Power and cooling: AI-factory design moves toward integrated systems
GIGABYTE and Schneider Electric announced a reference-architecture collaboration for NVIDIA Vera Rubin NVL72 systems, combining compute and networking with power, liquid cooling and facility management. It is a design effort, not a disclosed deployment or customer order. The economic significance is that coordinating facility systems with compute at the planning stage may reduce integration friction and help operators use scarce power and floor space more effectively as rack densities rise. ([blog.se.com](https://blog.se.com/datacenter/2026/11/13/gigabyte-and-schneider-electric-ai-factory-infrastructure-for-nvidia-vera-rubin-nvl72/))
Networking: Marvell spotlights interconnect and memory bottlenecks
Marvell says it will demonstrate optical interconnect, switching, CXL memory expansion and high-speed cable technologies at Supercomputing 2026. The announcement highlights a constraint beyond accelerator supply: compute can be underused when memory capacity or data movement cannot keep pace. If the demonstrated technologies translate into adoption, they could improve utilization of expensive AI clusters; the announcement itself does not establish customer orders or realized performance gains. ([marvell.com](https://www.marvell.com/company/newsroom/marvell-connectivity-ai-infrastructure-supercomputing-2026.html))
Robotics and applications: OpenAI reorganizes its robotics work
OpenAI confirmed to Fortune that it has dissolved its stand-alone robotics team, moving robotics researchers into its main research organization and applied robotics and hardware staff into its Applied organization. The company describes the change as a commitment to robotics, with more general-purpose models intended to control robots. Economically, this is a bet on reusing costly foundation-model research across physical tasks rather than maintaining a separate robotics stack. It may redirect investment and intensify pressure on robotics firms whose differentiation depends on specialized models; deployment economics remain unproven. ([fortune.com](https://fortune.com/2026/11/13/openai-disbands-robotics-team-as-its-general-models-start-running-robots/))
Robotics: GXO and Apptronik plan conditional warehouse trials
GXO and Apptronik signed a letter of intent to expand Apollo humanoid-robot trials, with trials planned to begin in 2027. They say a global rollout of up to 1,000 robots over four years is conditional on meeting agreed performance, safety and operational KPIs; initial North American deployment is targeted for late 2027 under a proposed Robotics-as-a-Service framework. For logistics operators, the model could shift spending toward service-based automation and address repetitive labor needs, but the scale-up is contingent—not a firm purchase commitment. ([gxologistics.gcs-web.com](https://gxologistics.gcs-web.com/node/11346/pdf))
Platforms and cyber: Fireworks AI breach exposes downstream customer risk
Fireworks AI confirmed a breach affecting customer billing information and datasets; TechCrunch reported that Notion said chat transcripts for some Notion AI users were accessed, and that around 1% of its customers were affected. The incident occurred in October but was reported within this update window. Its economic significance is the potential for a single AI-provider compromise to create downstream incident-response costs, customer churn and greater diligence demands across the model-provider supply chain. ([techcrunch.com](https://techcrunch.com/2026/11/12/cyberattack-on-fireworks-ai-allowed-hackers-to-access-notion-customers-chat-transcripts/?trk=article-ssr-frontend-pulse_little-text-block))
Several announcements describe plans, demonstrations or conditional trials rather than completed deployments or binding orders. The Fireworks incident occurred in October and was disclosed within this reporting window. Nvidia’s leadership transition is confirmed, but operational or financial effects are not yet established.