Amid a deepening structural decoupling between the US and China and the full-scale launch of the artificial intelligence (AI) investment cycle, the map of American imports is undergoing one of its most consequential shifts in decades.
At the latest earnings call on February 25, Nvidia chief executive Jensen Huang said deploying data centers in space is technically feasible but currently uneconomic. However, he expects conditions to improve over time as engineering approaches evolve and space-based computing use cases expand.
At the earnings call on February 25, Nvidia detailed its next-generation Rubin platform and Vera CPU roadmap as Blackwell systems continue to ramp, with management signaling sustained demand across Hopper, Blackwell, and forthcoming products while acknowledging uncertainty around the timing and scale of Rubin revenue.
Coretronic Corp. has accelerated its transformation from a traditional backlight module supplier into a diversified technology group spanning AI sensing, drones, and automated logistics. This strategic pivot is now delivering tangible results, with two key subsidiaries — Coretronic Intelligent Robotics Corp. (CIRC) and Coretronic MEMS Corp. (CMC) — returning to profitability in 2025.
Nvidia said uncertainty over China shipments and tight product supply remain key constraints, even as generative AI drives record capital spending by hyperscalers and sovereign customers accelerate national AI infrastructure investments.
The Trump administration has staked much of its AI credibility on Stargate — a US$500 billion infrastructure push announced alongside Sam Altman and Masayoshi Son. The project was meant to anchor a new era of American dominance in AI. More than a year on, it has yet to move beyond rhetoric.
Qualcomm has begun delivering rack-scale AI hardware and software systems for data centers, built around its AI 100 inference chip. The move signals a renewed push into a market where Nvidia and AMD currently set the standard.
During the earnings call on February 25, Nvidia said record capex plans by major cloud service providers reflect a structural shift toward monetizable AI workloads, with management arguing that token-driven revenue models support continued elevated infrastructure investment.
On February 25, Nvidia's blowout fourth-quarter results and bullish fiscal-2027 guidancehelped dispel recent market worries that the AI spending boom may be an unsustainable bubble, as the company reported record sales and signaled continued rapid demand for data‑center compute.
Nvidia is extending its enterprise infrastructure strategy into operational technology (OT) cybersecurity, applying accelerated computing and AI to protect energy grids, manufacturing plants, transportation networks, and utilities.
Anthropic has released a report accusing DeepSeek, Moonshot AI and MiniMax of systematically extracting capabilities from its Claude large language model through large-scale distillation. The goal: to accelerate training of their own systems.
More coverage
