Qwen3.8 Flash Next FP4 recipe
One-command vLLM recipe for Qwen3.8-Flash-Next-FP4-Local on a single DGX Spark, with an OpenAI-compatible API.
INDEX / 2026Writing & work
18 articles on local AI, Kubernetes, Linux, security and learning, plus 4 notes, 5 projects and 6 certifications. Articles and notes are local. The source projects are ported here; the two newest projects open on GitHub.
01 Archive
One DGX Spark, four ways to serve Qwen3.8-Flash-Next
I served Qwen3.8-Flash-Next four ways on one DGX Spark. Answer quality is close. Speed, memory and structured output are not.
Why I'm Betting on Local AI (and How You Can Start)
OpenAI just halved its $200 plan and closed models keep adding guardrails. Why I'm betting on local AI, and how to start with quantization, inference engines, recipes and harnesses.
DeepSeek V4 Flash vs GLM 5.3 Flash: Who Wins on a Single Spark?
DeepSeek V4 Flash faces GLM-5.3 Flash EXL3 on one DGX Spark across tool calling, serving, and 128k-context decode benchmarks, with full measured results.
DeepSeek V4 Flash on a Single DGX Spark
I ran DeepSeek V4 Flash on one DGX Spark with the MiaAI-Lab recipe, benchmarked throughput to 256k context, compared a GGUF path, and wired it into a Hermes kanban agent pipeline.
Running Laguna S 2.1 on a DGX Spark
How I fit and serve Poolside Laguna S 2.1 on a single NVIDIA DGX Spark using NVFP4, vLLM, DFlash speculative decoding, and a hardened thinking-off configuration.
My Obsidian PKM Setup: Karpathy's LLM Wiki + Claude Code + Local Search
My Obsidian PKM setup using Karpathy's LLM wiki pattern: Claude Code maintains the wiki, and qmd handles local search without the usual RAG overhead.
Gemma 4 E4B vs 26B on an RTX 4070 Ti: Benchmarks, RAG, and a Real Webapp Test
I benchmarked Gemma 4 E4B and Gemma 4 26B locally with llama.cpp on an RTX 4070 Ti to see which one is better for local RAG, web retrieval, and grounded summaries.
How I Passed KCNA and KCSA — And What to Actually Study
Passed KCNA and KCSA on the road to Kubeastronaut. Here's the exact courses, mock exams, and topics that matter — especially if you already hold CKA or CKAD.
Ghostty Terminal: Could This Be the Best Terminal Ever?
The time has finally come! Ghostty Terminal was officially released as version 1.0.0 at the end of December. I've been eagerly waiting for this fast, GPU-accelerated terminal emulator to…
Dagger and GitHub Actions: Simplifying CI/CD Pipelines
Back when I first discovered Jenkins, I was so amped up that, with a Jenkinsfile and a few steps, I could set up a full-fledged CI/CD pipeline for any application. But over time, that…
Arch Linux on VMware: Setting Up Sway and Wayland
This blog post is the second article in the series: "ArchLinux: The hard way". Since the release of the free version of Workstation pro 17, I wanted to try out the product and see how it…
Arch Linux installation with Encryption, LVM, GNOME on VirtualBox: The Hard way
I stumbled upon the fact that many tech enthusiasts I follow use Arch Linux as their daily driver, so I decided to give it a try. While waiting for my Framework laptop to arrive, I set up a…
Cross-Platform Dotfiles with Chezmoi, Nix, Brew, and Devpod
This guide walks you through the structure and approach I've taken to manage my productivity workflow with dotfiles. My setup is designed to work across both macOS and Linux distros,…
Simplifying Secret Management with Azure Key Vault and External Secret Operator
Secret handling has always been a pain point for me over the past few years. Most of the time, I ended up using services with features I didn't need, and even worse, wasting money on them.…
Homelab: Installation and Configuration of Talos on Raspberry Pi
Talos is rapidly becoming the go-to operating system for Kubernetes clusters, and for good reason. As part of my ongoing experiments with my homelab, I decided to see if the hype was…
Starting Small: The Beginning of My Kubernetes Homelab Journey
After years of anticipation, I've finally embarked on my homelab journey. The shift to remote work post-COVID has gifted me the time to pursue this long-awaited hobby. Gone are the days of…
Code, Clusters, and Gelato: A DevOps Summer Story
It's 3 PM on August 12th, and I'm sitting at my desk. Outside, the summer heat is intense, but my shutters are closed, creating a cocoon of coolness around me. Suddenly, my sister bursts…
How to be productive in the post covid era
Here's what every productivity article won't tell you: reading more productivity hacks makes you less productive. I spent two years collecting tips from Atomic Habits, Deep Work, and GTD,…
No articles in this topic yet.
02 Working notes
03 Projects
One-command vLLM recipe for Qwen3.8-Flash-Next-FP4-Local on a single DGX Spark, with an OpenAI-compatible API.
Local-first AI workspace for one DGX Spark: OCR, speech, translation, image and video generation, mind maps and chat on your own models.
DevPod dotfiles management using Chezmoi for consistent development environments across machines.
Personal Kubernetes homelab setup for self-hosting and infrastructure experimentation.
Command-line tool for managing and organizing notes efficiently from the terminal.
04 Certifications
Ported from devopsfortunato/alfonsofortunato.com at 526b447 on 19 Oct 2026. Writing and notes live here. Source projects are ported locally; newer GitHub projects link to their repositories.