corpus.blog
41554 blogs · [ { "id": "01a0d865-bff7-7210-9659-d8a2feb70f78", "title": "Laya vs Dijkstra: can a typed-decision model find its way out of a maze?", "url": "https://www.fratepietro.com/2026/laya-vs-dijkstra-path-solving-mlx/", "published_at": "2026-09-23T22:00:00+00:00" }, { "id": "01a0d865-bff7-7210-9659-d8a2ff49e0ed", "title": "LLM Inference: a learning roadmap", "url": "https://www.fratepietro.com/2026/llm-inference-learning-roadmap/", "published_at": "2026-09-22T22:00:00+00:00" }, { "id": "01a0c50d-b54a-7028-864c-1980aee8e769", "title": "Bonsai: a 27B reasoning model on a 16 GB M2 Mac, with Frink", "url": "https://www.fratepietro.com/2026/frink-bonsai-ternary-27b-local/", "published_at": "2026-09-17T22:00:00+00:00" }, { "id": "01a0b0d0-a384-7272-ab01-729c82b0b9c7", "title": "Cognitora v0.9.1: gossip discovery, live KV prefix, and a native engine", "url": "https://www.fratepietro.com/2026/cognitora-inference-update-v0.9/", "published_at": "2026-09-14T22:00:00+00:00" }, { "id": "01a08779-c879-7007-8a6a-8349b661fa13", "title": "gitgui: a real git GUI inside cmux, next to Pi", "url": "https://www.fratepietro.com/2026/gitgui-terminal-git-gui-cmux-pi/", "published_at": "2026-09-02T22:00:00+00:00" }, { "id": "01a08779-c879-7007-8a6a-8349b723b4af", "title": "Ferrox on Metal: at parity with llama.cpp, and past it", "url": "https://www.fratepietro.com/2026/ferrox-metal-parity-llama-cpp/", "published_at": "2026-08-23T22:00:00+00:00" }, { "id": "01a0c50d-b54a-7028-864c-1980b063b406", "title": "Frink on Metal: at parity with llama.cpp, and past it", "url": "https://www.fratepietro.com/2026/frink-metal-parity-llama-cpp/", "published_at": "2026-08-23T22:00:00+00:00" }, { "id": "01a08779-c879-7007-8a6a-8349b7a5780f", "title": "Ferrox: a Rust GGUF engine, measured against llama.cpp", "url": "https://www.fratepietro.com/2026/ferrox-rust-gguf-inference-engine/", "published_at": "2026-08-04T22:00:00+00:00" }, { "id": "01a0c50d-b54a-7028-864c-1980b0a08817", "title": "Frink: a Rust GGUF engine, measured against llama.cpp", "url": "https://www.fratepietro.com/2026/frink-rust-gguf-inference-engine/", "published_at": "2026-08-04T22:00:00+00:00" }, { "id": "01a08779-c879-7007-8a6a-8349b7b970f2", "title": "Rondine: Hardware-Aware Local LLM Serving Without the Flag Archaeology", "url": "https://www.fratepietro.com/2026/rondine-hardware-aware-local-llm-launcher/", "published_at": "2026-07-21T22:00:00+00:00" }, { "id": "01a08779-c879-7007-8a6a-8349b7ca962c", "title": "Running GLM-5.2 Locally with Rondine and Pi", "url": "https://www.fratepietro.com/2026/running-glm-5-2-locally-rondine-pi/", "published_at": "2026-07-21T22:00:00+00:00" }, { "id": "01a08779-c879-7007-8a6a-8349b7edd14c", "title": "The Local Agent Trio: cmux + Pi + Unsloth Studio", "url": "https://www.fratepietro.com/2026/cmux-pi-unsloth-local-glm-setup/", "published_at": "2026-06-28T22:00:00+00:00" }, { "id": "01a08779-c879-7007-8a6a-8349b8b4411c", "title": "Google AX: A Distributed Agent Runtime with Durable Execution", "url": "https://www.fratepietro.com/2026/google-ax-distributed-agent-runtime/", "published_at": "2026-06-04T22:00:00+00:00" }, { "id": "01a08779-c879-7007-8a6a-8349b8f5408c", "title": "Agent Substrate: Zero-Idle Kubernetes for Stateful AI Agents", "url": "https://www.fratepietro.com/2026/agent-substrate-zero-idle-kubernetes/", "published_at": "2026-06-02T22:00:00+00:00" }, { "id": "01a08779-c879-7007-8a6a-8349b9f10c72", "title": "DwarfStar 4: antirez Bets the Farm on Local Inference Done Right", "url": "https://www.fratepietro.com/2026/dwarfstar-4-local-inference-antirez/", "published_at": "2026-05-29T22:00:00+00:00" }, { "id": "01a08779-c879-7007-8a6a-8349badd4d7c", "title": "Cognitora: A Datacenter-Scale, Open-Source LLM Inference Orchestrator (NVIDIA Dynamo Alternative)", "url": "https://www.fratepietro.com/2026/cognitora-inference-llm-orchestration/", "published_at": "2026-05-02T22:00:00+00:00" } ] posts Claim your blog
Back to fratepietro.com
Blog · corpus.blog/blogs/fratepietro.com/posts

fratepietro.com

fratepietro.com

2026

Laya vs Dijkstra: can a typed-decision model find its way out of a maze?original ↗
23 Sept 2026
LLM Inference: a learning roadmaporiginal ↗
22 Sept 2026
Bonsai: a 27B reasoning model on a 16 GB M2 Mac, with Frinkoriginal ↗
17 Sept 2026
Cognitora v0.9.1: gossip discovery, live KV prefix, and a native engineoriginal ↗
14 Sept 2026
gitgui: a real git GUI inside cmux, next to Pioriginal ↗
2 Sept 2026
Ferrox on Metal: at parity with llama.cpp, and past itoriginal ↗
23 Aug 2026
Frink on Metal: at parity with llama.cpp, and past itoriginal ↗
23 Aug 2026
Ferrox: a Rust GGUF engine, measured against llama.cpporiginal ↗
4 Aug 2026
Frink: a Rust GGUF engine, measured against llama.cpporiginal ↗
4 Aug 2026
Rondine: Hardware-Aware Local LLM Serving Without the Flag Archaeologyoriginal ↗
21 Jul 2026
Running GLM-5.2 Locally with Rondine and Pioriginal ↗
21 Jul 2026
The Local Agent Trio: cmux + Pi + Unsloth Studiooriginal ↗
28 Jun 2026
Google AX: A Distributed Agent Runtime with Durable Executionoriginal ↗
4 Jun 2026
Agent Substrate: Zero-Idle Kubernetes for Stateful AI Agentsoriginal ↗
2 Jun 2026
DwarfStar 4: antirez Bets the Farm on Local Inference Done Rightoriginal ↗
29 May 2026
Cognitora: A Datacenter-Scale, Open-Source LLM Inference Orchestrator (NVIDIA Dynamo Alternative)original ↗
2 May 2026