8-core CPUs and 16GB VRAM become the most common Steam configurations
Memory management signal - matters for fitting large models across 72GB total VRAM.
Independent publication // local-first AI // field reports from owned systems
Nodehome
A running publication about local models, private inference, self-hosted agents, weird hardware, research sweeps, and the builders wiring their own AI stack together.
AI is getting physical again. It shows up in terminals, racks, side projects, and ugly little workflows people actually control.
Latest
Daily Sweep - Aug 03, 2026Chinese company puts AMD B650 chipset on a PCIe storage card for $90
Bus or storage bandwidth topic - relevant to the H12SSL-i's PCIe 4.0 layout.
AirLLM 70B inference with single 4GB GPU
Sakana Namazu: An LLM API with Japanese-vibes! 🎏 Built for Japanese enterprises, featuring frontier-level reasoning and built-in agentic tools. Blog post: sakana.ai/namazu-api#E...
Agent or tool-use pattern - relevant to local AI workflow automation.
Show HN: Kakehashi - Experimental userspace to run macOS binaries on Linux ARM
AMD Radeon GPUs are getting per-chip VRAM temperature monitoring in HWiNFO
Memory management signal - matters for fitting large models across 72GB total VRAM.
Field Reports
builds, experiments, notesThree RTX 3090s, One 32B Model: A Pipeline-Parallel Canary
A current field note on why the 3x3090 serving path moved through pipeline parallelism, not tensor parallelism, for the tested 32B AWQ model.
Gemma 4 12B And The Sensory Agent Lane
A public-safe read on Gemma 4 12B as a local sensory preprocessor: useful for seeing, hearing, and structuring observations without turning into an action system.
Hardware
machines, thermals, economicsPower Caps On Three RTX 3090s: Bursts Versus Sustained Load
A measured note on 300W bursty inference, lower caps for sustained runs, and why power-cap sweet spots are workload-specific.
Parallel Agent Serving Is A Hardware Shape Now
A field-report read on 14x RTX 3090 agent serving, EXL3, FP8 KV cache, Aphrodite, and why concurrency is becoming the local hardware metric.