Nvidia Dgx Spark
- Source
- factory.com
- Author
- Factory News
- Date
Context
Factory describes a deployment path that connects its Droid software-engineering to an approved local model on NVIDIA's DGX Spark, a system it presents as a local alternative to data center compute. The announcement names Nemotron 3.5 Lightning and other locally deployable models as options. Keeping model computation inside the team's environment addresses a constraint for organizations handling sensitive source code and operational information.
The intended workflow still ends with an engineer reviewing the proposed change: the agent investigates code, modifies it, and runs tests. The article directs organizations seeking on-premises or air-gapped installations to contact Factory for the appropriate installation path, rather than presenting those configurations as automatic consequences of using the hardware.
- AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
- mixture-of-experts — A model built from many specialist sub-networks where only a few activate per token, giving big-model capability at small-model running cost.
Checking sign-in…
Loading comments…