A Panther Lake Mini PC: The Asus Nuc Pro 16 Benchmark and Review
Level1Techs - deep-dive hardware and Linux content for builder types.
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 - Jul 22, 2026I can prove I had the best time at Open Sauce
Geerling content - practical hardware testing, often with Linux and ARM angles.
Jack Dorsey launches Buzz to combine team chat, AI agents and Git hosting
Agent or tool-use pattern - relevant to local AI workflow automation.
The State of Simulation for Physical AI: An Overview
Hugging Face blog post - check for new model releases, library updates, or ecosystem shifts.
Open Sauce and GPS time were my summer AI Antiseptics
Geerling content - practical hardware testing, often with Linux and ARM angles.
Incredibly proud of the Sakana AI team. We have developed an orchestration model right here out of Japan that achieves state-of-the-art performance on real-world cybersecurity benchmarks! 🎌 sakana.ai/fugu-cyber-r...
Performance data that could inform model selection or serving config.
Judge approves $1.5B Anthropic settlement for pirated books used to train Claude
Cloud/closed-model news - context for the local-first value proposition.
NVIDIA’s first GeForce driver for Windows on Arm confirms RTX Spark N1X with 6,144 or 5,120 CUDA cores
GPU compute backend change - check if it affects CUDA serving on the 3090s.
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.