NVIDIA AI News - Page 10 of 28

Seven decision models just went up against a generative LLM baseline, and the gap in both speed and accuracy is wide enough to notice.

Donald Trump and Xi Jinping sit down Thursday for a state dinner at the White House, and the guest list tells its own story: OpenAI and Nvidia...

Mapping a heterogeneous permeability field to a pressure distribution , that’s the core of Darcy flow, a fundamental problem in subsurface modeling...

Behind the staggering $20 billion acquisition of Groq lies a quiet admission: Nvidia’s GPUs are not built for every kind of AI inference.

Servers built around Nvidia's Vera Rubin and Grace Blackwell chips are getting more expensive, and by a lot.

Nvidia's stock market story runs through AI data centers, but the company is spending heavily on a different bet: machines that move through the...

South Korea wants to buy a quarter of a million rocket boosters for its AI program.

Z.ai released GLM-5.3 this week, and the number that stands out isn't a benchmark score. It's zero: the company didn't retrain its base model.

Forty million new supervised and alignment samples. That’s the raw fuel behind Nemotron 3 Super, not just more data, but a deliberate architecture of...

An engineering team just ported 24 GPU operators, roughly 40 kernels in total, from Python to Rust without a human rewriting a single one by hand.

Virtual screening runs against millions to billions of compounds, and co-folding models like OpenFold3 often produce the most accurate structures in...

Training today's massive AI models creates a monumental traffic jam. The gridlock isn't the raw horsepower of the GPUs themselves.

The ONNX graph is a labyrinth of nodes, each one a decision point. But when you’re chasing FP8 performance, the path isn’t just about layout, it’s...

The data center is no longer a room of humming servers, it’s a factory. And NVIDIA is engineering its blueprint from the ground up.

Waldron called it “viral.” JPMorgan’s chief analytics officer didn’t expect employees to build their own AI assistants, complete with personas,...

Jina AI, the Berlin-based search and embeddings company now under Elastic, released jina-ocr-v1 this week, a document parser built to run on GPUs...

It took 149 years, from the camera's invention in 1826 to 1975, for humans to produce 1.5 billion images.

The race to scale long-context AI has a new contender. Moonshot AI just dropped FlashKDA, a production-grade CUDA kernel that turns their Kimi Delta...

Training large language models at 4-bit precision without sacrificing accuracy has long seemed like a zero-sum game, until now.

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