AI Hardware News - GPUs, TPUs, Chips & Data Center Updates

KernelEvolve: How Meta’s Ranking Engineer Agent Optimizes AI Infrastructure

This is the second post in the Ranking Engineer Agent blog series exploring the autonomous AI capabilities accelerating Meta’s Ads Ranking innovation. The previous post introduced Ranking Engineer Agent’s ML exploration capability, which autonomously designs, executes, and analyzes ranking model experiments. This post covers how to optimize the low-level infrastructure that makes those models run [...] Read More... The post KernelEvolve: How Meta’s Ranking Engineer Agent Optimizes AI Infrastructure appeared first on Engineering at Meta .

Scaling seismic foundation models on AWS: Distributed training with Amazon SageMaker HyperPod and expanding context windows

This post describes how TGS achieved near-linear scaling for distributed training and expanded context windows for their Vision Transformer-based SFM using Amazon SageMaker HyperPod. This joint solution cut training time from 6 months to just 5 days while enabling analysis of seismic volumes larger than previously possible.

Control which domains your AI agents can access

In this post, we show you how to configure AWS Network Firewall to restrict AgentCore resources to an allowlist of approved internet domains. This post focuses on domain-level filtering using SNI inspection — the first layer of a defense-in-depth approach.

Brain-inspired chip could make some AI tasks up to 2,000 times more energy efficient

A new type of computer chip that uses the physics of materials to process information could make some artificial intelligence (AI) systems far more energy efficient, researchers have found. Loughborough University physicists have developed a device that can process data that changes over time directly in hardware, rather than relying on software running on conventional computers.

Meta’s natural gas binge could power South Dakota

Meta's upcoming Hyperion AI data center will be powered by 10 new natural gas plants.

Cognichip wants AI to design the chips that power AI, and just raised $60M to try

The firm says it can can reduce the cost of chip development by more than 75% and cut the timeline by more than half.

Chip-scale light technology could power faster AI and data center communications

Researchers at Trinity have developed a new light-based technology on a tiny chip that could help make the data centers behind cloud computing, artificial intelligence, and global internet services faster and more efficient. In the new research, recently published in Nature Communications, the Trinity team reported one such promising advance with collaborators at the University of Bath and the Swiss Federal Institute of Technology Lausanne (EPFL).

Meta Adaptive Ranking Model: Bending the Inference Scaling Curve to Serve LLM-Scale Models for Ads

Meta continues to lead the industry in utilizing groundbreaking AI Recommendation Systems (RecSys) to deliver better experiences for people, and better results for advertisers. To reach the next frontier of performance, we are scaling Meta’s Ads Recommender runtime models to LLM-scale & complexity to further a deeper understanding of people’s interests and intent. This increase [...] Read More... The post Meta Adaptive Ranking Model: Bending the Inference Scaling Curve to Serve LLM-Scale Models for Ads appeared first on Engineering at Meta .

Building sovereign AI at the edge: Microsoft and Armada collaborate to deliver Azure Local on Galleon modular datacenters

Explore how Microsoft and Armada bring sovereign AI to the edge with Azure Local, enabling secure, resilient workloads in disconnected and regulated environments. The post Building sovereign AI at the edge: Microsoft and Armada collaborate to deliver Azure Local on Galleon modular datacenters appeared first on Microsoft Azure Blog .

ScaleOps raises $130M to improve computing efficiency amid AI demand

ScaleOps just raised $130M to tackle GPU shortages and soaring AI cloud costs by automating infrastructure in real time.

AI chip startup Rebellions raises $400 million at $2.3B valuation in pre-IPO round

The startup, which is planning to go public later this year, designs chips specifically for AI inference, another challenger to Nvidia's dominance.

Mistral AI raises $830M in debt to set up a data center near Paris

Mistral aims to start operating the data center by the second quarter of 2026.

Microsoft takes over a Texas AI data center expansion after OpenAI backs away

Microsoft is taking over a data center construction project in Texas after OpenAI declined to pursue it, in a move that will make the two companies neighbors at one of the nation's largest complexes for running artificial intelligence.

[AINews] H100 prices are melting *UP*

a quiet day lets us report an important GPU trend

Claude rate limit chaos

VCs are betting billions on AI’s next wave, so why is OpenAI killing Sora?

When an 82-year-old Kentucky woman was offered $26 million from an AI company that wanted to build a data center on her land, she said no. Sure, that same company can try to rezone 2,000 acres nearby anyway, but as AI infrastructure stretches further into the real world, the real world is starting to push back. That tension is everywhere […]

AI's arrival complicates Big Tech climate goals, and some worry it's locking in more fossil fuels

Six years ago, Google was confident that by 2030 it would power all operations with electricity generated from clean sources, including wind and solar power, and remove as much pollution as it produced. Today it calls those goals a moonshot. Microsoft says it's still aiming to remove more carbon than it creates by 2030 but now describes the effort as a marathon, not a sprint.

Nvidia's new AI tool is giving female game characters a makeover—and gamers are pushing back

Last week leading chipmaker Nvidia announced DLSS-5 (Deep Learning Super Sampling), a new artificial intelligence (AI) rendering tool it describes as a breakthrough in visual fidelity for games. The software takes low-resolution images and uses AI to upscale them, adding what Nvidia calls photoreal lighting and materials.

Brain-inspired AI hardware helps autonomous devices operate efficiently and independently

The human brain constantly makes decisions. It requires minimal power to move bodies in a desired direction or avoid an object. A Purdue University engineer uses the brain's efficiency as inspiration to help autonomous vehicles, such as drones and robots, make crucial, time-sensitive decisions while operating in the field.

OpenAI shuts down Sora while Meta gets shut out in court

When an 82-year-old Kentucky woman was offered $26 million from an AI company that wanted to build a data center on her land, she said no. Sure, that same company can try to rezone 2,000 acres nearby anyway, but as AI infrastructure stretches further into the real world, the real world is starting to push back. That tension is everywhere […]