Urgent.News

What's breaking now, across thousands of outlets.

AI

Alibaba Cloud plans six-year stroll to 20GW of datacenters, reveals chip to power them

USA alone has 37GW under construction, but Alibaba doesn't think the AI boom will run out of steam

Alibaba Cloud plans six-year stroll to 20GW of datacenters, reveals chip to power them

Alibaba, the Chinese tech giant, has announced plans to expand its datacenter fleet to 20 gigawatts of capacity over the next six years. At the Apsara conference, CEO Eddie Wu likened early AI applications to light bulbs, but emphasized that AI will eventually drive 99.9% of all cognition, much like steam and combustion engines revolutionized physical work after their initial introduction.

Wu compared AI to steam engines, explaining that just as engines have become the primary source of physical labor, AI will eventually handle most cognition tasks. He hinted at the potential for AI to create millions of "AI scientists and domain experts" constantly pushing the boundaries of technology. Alibaba Cloud plans to build its share of this infrastructure by scaling its global datacenter fleet to 20GW, with a large portion going towards the Stargate project, which aims to bring 10GW online by 2029.

Alibaba's own Zhenwu V900 chip, described as the most powerful AI chip in China today, will be utilized for frontier model training and inference in its giant clusters.

Written by urgent.news from The Register's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Also reported by 1 other outlet

Read the original at theregister.com →

More in AI

Observability in Microsoft Foundry: Tracing Agent Runs, Continuous Evaluation, and the OpenTelemetry Data Plane

Day 10 of the Microsoft Foundry 100 Days / 100 Blogs series. You shipped an agent. It calls a model, invokes two tools, retrieves a few documents, and returns an answer. It works in your dev loop.

  • Microsoft Foundry uses OpenTelemetry for distributed tracing, metrics, and logs.
  • Traces represent a single request's journey through the system, identified by a traceid.
  • Foundry exports traces to Azure Monitor Application Insights for unified observability.

5 Guardrails That Keep an LLM Agent Shippable in Production

An LLM agent that demos well and an LLM agent that survives production are almost different projects. The demo needs the happy path. Production needs everything else.

  • Limit agent to small set of tools with typed inputs and server-side validation
  • Place human in the loop for high-stakes actions, establish clear threshold
  • Build evaluation harness with real cases for every change

More from Tuesday 22 September →