#Hardware

Arm prepares its first in-house chips for Meta as SpaceX lifts space stocks

Keir Starmer's AI bet in Britain has raised questions about the promised billions

Russian scientists propose cooling for AI data centers that cuts electricity use by up to 22%

Hardware·Habr AI

OpenClaw turned out to be a costly autonomous agent: 81 million tokens and server conflicts

Nvidia Nemotron-Cascade-2 was run at home on a GeForce RTX 3090 at up to 150 tokens/s

Meta to build seven gas-fired power plants for new AI data center in Louisiana

U.S. Congress proposes tightening chip equipment shipments to China through allies

A bipartisan bill introduced in the U.S. would tighten shipments of chipmaking equipment to China, including exports from the Netherlands an

01Huawei's growth slowed after Apple's surge in China and intensifying competition in AI chipsMay 2, 2026·3 min 02Jensen Huang showed how Nvidia is redefining agentic AI infrastructureMay 2, 2026·2 min 03TSMC to further prioritize AI orders amid a 3nm chip shortage by the end of the third quarter of 2026May 2, 2026·2 min 04Fujitsu to have Rapidus produce a 1.4 nm AI chip, aims to integrate it into servers by 2030May 2, 2026·2 min 05Alphabet nearly caught up with Nvidia by market cap after Google Cloud's record quarterMay 2, 2026·3 min

Top themes

Microsoft and Chevron discuss a $7 billion energy complex for data centers in Texas

Microsoft is in exclusive talks with Chevron and Engine No. 1 on a long-term agreement that would support the construction of an energy comp

01Valar Atomics raises $450 million for small nuclear reactors for AI data centersMay 2, 2026·3 min 02Samsung and SK Hynix move AI customers to 3–5-year contracts with a price floorMay 2, 2026·2 min 03Microsoft CFO Amy Hood: betting on AI amid fears of a new bubbleMay 2, 2026·3 min

Hardware·3DNews AI

Ollama accelerates local AI on Apple M5: a Mac with at least 32 GB of memory is required

Ollama 0.19 adds an MLX backend for Apple M5, M5 Pro and M5 Max: local models and code agents respond faster, but a Mac with 32 GB of unifie

Hardware·Habr AI

Repka-Pi 4 gets local text-to-speech with Piper and FastAPI, without a GPU and with autostart

On Repka-Pi 4, a local TTS stack with eSpeak NG and Piper was demonstrated: neural speech synthesis works without a GPU, accepts HTTP reques

Hardware·Hugging Face Blog

Hugging Face added gradio.Server: custom frontends can now connect to a Gradio backend

Hugging Face’s new gradio.Server turns Gradio into a backend layer for React, Svelte, and plain HTML/JS, while preserving request queues, ZeroGPU, and compatibility with Spaces.

Hardware·CNews AI

Russians are building home supercomputers en masse to work with neural networks

Amazon may sell its AI chips to third parties as the business heads toward $20 billion

Hardware·TNW

TSMC increased quarterly revenue by 35% and confirmed resilient demand for AI chips

PJM seeks 15 gigawatts of new generation capacity to meet AI data-center demand

Arm changes course: smartphone chip architect moves into AI, clouds, and data centers

Arm changes course: René Haas's company bets on data centers and AI

Arm, long known as a supplier of smartphone architectures, is shifting its strategy toward the cloud, data centers, and generative AI — wher

Hardware·Habr AI

STMicroelectronics' STM32N6 demonstrated on-device speech recognition without the cloud at 0.2 W

An experiment on the STM32N6 with an integrated NPU showed that the microcontroller can recognize free-form speech locally, almost in real t