Source

NVIDIA Blog i NVIDIA Newsroom

2 items

NVIDIA's own corporate blog, newsroom and product pages. Primary for what the company itself announces: acquisitions, prices, hardware dates and the software it ships. Its performance figures are the vendor's own measurement and need attribution; the technical detail behind them usually sits on developer.nvidia.com, which is a separate record.

blogs.nvidia.com ↗

Nvidia's artwork for the Nemotron 3.5 release: a row of dark tiles with icons for search, code, charts and video, around a glowing green cube, in front of the Earth seen from space.

Machines & local models

Nemotron 3.5 Bolt squeezes 74 billion parameters into 24 GB, but Nvidia's llama.cpp recipes start at 32 GB

Nvidia released two open models for agents on October 15, 2026: the large Nemotron 3.5 Super VL and the smaller Nemotron 3.5 Bolt 74B-A9B, built to run on your own machine. Two-bit quantization brings Bolt down to a 24 GB package. Nvidia's own llama.cpp instructions, though, call for a GPU with at least 32 GB or an RTX Spark with 64 GB of memory.

NVIDIAOctober 19verified

Agents

Nvidia ships an open-source sandbox for agents and a hardware watchdog that quarantines them

Nvidia announced the Open Agent Safety Platform on September 28, 2026. It pairs OpenShell, an open-source runtime under the Apache 2.0 license that confines an agent's file, process, and network access, with Sentry, a watchdog that runs on BlueField-4 data processing units and, according to Nvidia, stops an agent that leaves its boundary within milliseconds. Anthropic, Microsoft, Hugging Face, Salesforce, SAP, and OpenClaw are among the organizations Nvidia lists.

NVIDIA NewsroomSeptember 30verified