News

Tools & Products

Launches, features and updates you can use today.

844 stories · Every headline links to the original publisher. HapnGo curates and groups; we do not copy articles.

Tools & Products NVIDIA Technical Blog

Run AI Models for Edge Application Development with Just a Few Commands

AI summary NVIDIA has detailed how its TensorRT Edge-LLM, a lightweight C++ inference runtime, lets embedded developers prototype AI models on the new 64GB version of DGX Spark and later deploy them on Jetson Thor and other Thor platforms. A command-line tool called trtedge fetches a model, builds optimized engines and can start a chat with Nemotron-3.5 Lightning, whose first download is about 21GB. The model can also be served locally through an OpenAI-compatible API. With DSpark speculative decoding using three draft tokens, average generation on HumanEval prompts rose from 73.4 to 126.9 tokens per second, with peak memory at 27.9GB.

Tools & Products Product Hunt

Agnos

Clean documents for AI, and see what was removed Discussion | Link

Free, daily

The HapnGo Briefing, in your inbox

One email a day with the AI stories that matter, why they matter, and one tool worth trying. No spam, one-click unsubscribe.

We only use your address to send the briefing and, if you vote, to count one vote per reader. Privacy.