Forward Ollama from an NVIDIA DGX Spark to your laptop over SSH
aestheticedwar1 · x · 2026-10-14
The author shows how to run Ollama on an NVIDIA DGX Spark and forward port 11434 over SSH (ssh -L 11434:localhost:11434), so laptop tools talk to localhost as if the model were local while it actually runs on the Spark's 128 GB unified memory. A follow-up notes NetBird can replace port forwarding entirely, with zero inbound ports opened.
Related event: SSH Port Forwarding Exposes DGX Spark's Ollama as a Local Service(2 posts)→
More from Infra
- Runway partners with Baseten for dedicated enterprise deployment of Gen-4.5 and Aleph 2.0 — runwayml · 2026-10-14
- Data center panel asks 'Are We Screwed?' as Gary Marcus says AI industry may have screwed itself — GaryMarcus · 2026-10-14
- Australia drafts national AI and data center framework for 2027 as foreign money flows in — YvesMulkers · 2026-10-14
- ASML Posts €11.7B Q3 Net Sales, 57.5% Gross Margin — BenBajarin · 2026-10-14
- Repeated-Data Pretraining Study: Smaller Batches Act as Regularizers and Beat More GPUs — Sewon Min · 2026-10-14
- AWS walks through deploying and benchmarking models on SageMaker JumpStart with reusable Python script — AWS ML Blog · 2026-10-14