NVIDIA AI News - Page 5 of 28

Apple's on-device Dictation feature has a memory problem that has nothing to do with the language model everyone worries about.

Mark Zuckerberg is about to spend a breathtaking sum of money. Meta's capital expenditure is projected to blast past $100 billion in 2026.

Oracle is pitching its cloud like a utility company. Its new Zettascale10 infrastructure promises multi-gigawatt capacity—a unit for power grids, not...

Sarah Guo carries a handbag that happens to contain a robot brain. The founder of Conviction, an AI-focused venture capital firm, and co-host of the...

Speed. Precision. Scale. Step 3.7 Flash is no longer just a promising model , it’s a GPU-native powerhouse.

The Pentagon just locked in its roster. Eight companies—SpaceX, Oracle, and six other tech titans—now hold contracts to weave artificial intelligence...

Liquid AI put a number on its new vision model Tuesday: 35 milliseconds to process an image on the NVIDIA Jetson stack, fast enough for a factory...

AMD announced Wednesday it will invest up to $5 billion in Anthropic, tying the AI company's future computing needs directly to AMD's chip roadmap.

NVIDIA's Nemotron-3-Super has a familiar problem: it's accurate, but it's a hybrid Mamba-Transformer MoE with 120.7B total and 12.8B active...

Nvidia is buying Hugging Face for $12.9 billion, according to The Information, a price tag that works out to roughly 80 times the open-source AI...

Google Cloud will rent out its newest AI chip within weeks. The Ironwood TPU is a direct shot at Nvidia.

Eighty-nine percent of telecom companies now say open source models and software matter to their AI strategy, according to NVIDIA's latest State of...

NVIDIA is adding a fifth category to its AI networking stack, and this one points inward rather than out.

The AI agent gold rush is entering its industrial age. LangChain, the company whose open-source frameworks have quietly become the operating system...

Microsoft just fired its biggest cloud cannon at the competition. Bringing Anthropic's Claude models to Azure is that declaration of war.

The key-value cache is the lifeblood of modern AI inference, fast, fluid, and fiercely hungry for memory.

Your RAG evaluation scores are probably going up. This is not necessarily good news.

Perplexity started shipping Portable Computer this week, a local-first version of its agentic Computer platform built to run on NVIDIA's DGX Spark...

TensorRT deployments fail in predictable ways. One unsupported operation is all it takes.

DeepSeek closed a $7 billion funding round in late May at a $52 billion valuation.