Fathom Layer / Guides
Choosing hardware, deploying a model or connecting a system starts with constraints. Find the guide that addresses the question in front of you.
Read our editorial method ↗
Routes reflect each guide's title and introduction.
They are navigation aids, not quality rankings.
16 guides in this selection
A "swarm" is usually the wrong mental model. What works in production is a small, fixed set of specialised agents with defined hand-offs — a pipeline, not a crowd. CrewAI gives you the agent/task/crew structure; MCP gives every agent the…
Local Large Language Model (LLM) inference throughput is constrained by memory bandwidth (B/s) during the token generation phase and total VRAM capacity during model allocation. While modern enterprise GPUs like the Nvidia H100 offer…
The pattern: wrap your vector store in an MCP server so any MCP-aware client — an IDE, an agent, a chat app — can retrieve from it without a custom integration. Your documents never leave your machine; the client just calls a search tool.
An edge fleet has no perimeter to defend — the devices are in warehouses, vehicles, homes and factory floors, on networks you don't control. Zero trust is the only model that fits: every device, request and data flow is authenticated and…
WebNN lets a web page run neural-network inference on the device's NPU, GPU or CPU through the OS's own ML runtime — no model download of a runtime, no server round-trip, data never leaves the browser. In a Next.js app it is a client-only…
First, a correction the old guides got wrong: GGUF is an inference format, not a training format. You run GGUF models (llama.cpp, LM Studio, Ollama); you fine-tune in PyTorch and then convert. This guide covers both paths on Windows.
MCP gives an agent a uniform way to reach tools and data. That same uniformity is the attack surface: one malicious or compromised server sits inside the agent's trust boundary. These are the classes to design against.
Running a model locally with Ollama, llama.cpp or vLLM is a few commands. Exposing that server safely is the part the quickstarts skip. The failure mode is consistent: a port bound to 0.0.0.0, no authentication, reachable from the internet.
The question stopped being "which model?" and became "which engine runs it best on my hardware?". Three names cover most of the decision. Here is how they differ in the way that matters — without benchmark numbers that will not transfer to…
Local inference is memory-bandwidth bound, not compute bound. Generating one token requires streaming every active weight from memory exactly once. So the ceiling is:
Both are strong open-weight families you can run locally. The split is simple: DeepSeek's reasoning models think before answering (a visible chain-of-thought pass that costs tokens and latency), while Llama 3 answers directly. Pick based on…
The Tesla P40 is the cheapest route to a large VRAM pool: 24 GB per card on the used market for a fraction of a consumer 24 GB GPU. Two of them give you 48 GB for the price of one modern card. The catch is that it is a 2016 Pascal…
Most "wait for the next one" advice is wrong, because the next one is always six months away and the one after that is six months past it. The useful question is narrower: is a specific, dated improvement coming that changes your decision?
NPUs are fixed-function matrix execution blocks engineered to run low-precision tensor operations inside a 2W to 6W thermal envelope, preventing high-power SIMD vector lanes and discrete GPUs from depleting battery reserves. Paying a cash…
The honest answer is a range, and the number that sets it is not compute — it is memory. A model runs on your machine only if its weights, its KV cache and a working buffer fit in one pool of fast memory. Miss that by a gigabyte and…
An MCP server runs with whatever access you grant it and can read data, call APIs and — if you let it — take irreversible actions. Registries list servers; they do not vet them. Treat every server as untrusted third-party code, because that…
Use the tools to model assumptions, compare published products or clarify a term before making your choice.
