Paris Marx, author of Hyperscale, has followed data center development worldwide for years. He argues that US opponents have much to learn from experiences in places ranging from Ireland to New Zealand.
The US Treasury has proposed rules that would expand its Outbound Investment Security Program, widening limits on American investment in chips, AI and biotech in countries it considers adversaries. Treasury Secretary Scott Bessent said the aim is to stop US money from funding technologies that threaten national security.
Google has open-sourced ML Drift, a cross-platform GPU compute engine for on-device AI and machine learning inference, under Apache 2.0. It hides the differences between GPU hardware and low-level APIs and supports OpenGL ES, OpenCL, Metal and WebGPU.
Big tech companies are taking on ever more debt to fund vast AI infrastructure. The gap between the billions spent and the revenue expected is testing investor confidence.
At PyTorch Conference North America 2026, Google and Meta released TorchTPU, a native PyTorch backend for Google's TPUs, and Meta said it plans to shift larger LLM workloads onto TPUs with it. AWS released TorchNeuron for Trainium the same week, while developers questioned TPU availability and where vendor lock-in now lies.
Driven by artificial intelligence, the data center market is booming, and worldwide spending is expected to exceed $2 trillion in 2028.