RSSAmplifier

Blog

Mike Ayles

Trucks, engines, emissions, and the software around them.

mikeayles.comRSS feed ↗34 posts

Latest posts

Taalas-Style On-Chip Weights on a $250 FPGA: a Language Model at 60k tok/s

A 3.16M-parameter INT4 transformer running entirely in the on-chip memory of a Xilinx Kria KV260. Zero DRAM in the token loop, 59,965 tok/s on the fabric, bit-exact. Chat with it live.

On-Chip LLM: Sampling Without Asking

The FPGA decoded a reply in 6.6ms. It landed in the browser in 100. The gap was one unmeasured loop, and the fix was the Gumbel-max trick.

On-Chip LLM: Inside the Chip

The datapath of a 3.16M-parameter transformer living in FPGA fabric: wide-word GEMV, the dual-port split-brain, and the proof that 16 streams is a hard wall.

On-Chip LLM: Serving an FPGA to Strangers

Browser to transistor: a Cloudflare tunnel, a serving box, a Tailscale link, and the trick that lands a reply the instant you press Enter.

On-Chip LLM: War Stories and the Method

Bit-honest before fast. Where iverilog lies, why silicon beats the timing report, and the optimisation that won and got stopped anyway.

Are you sure OneDrive has got your back(up)?

OneDrive has a huge, blind hole if you're using it in/as your backup strategy. It silently fails if you have too many files.

Banger: A Local Photo Culler That Doesn't Touch the Cloud

A files-only photo culler for the laptop. CLIP embeddings, MMR diversity, ArcFace identity clustering, and a Ridge regressor trained on your own taste. Copies keepers out, never edits the originals.

Does RAG Actually Help AI Coding Tools?

I benchmarked Claude Code and GitHub Copilot on the same model with and without RAG-powered semantic search across 60 queries. Both tools search equally well — but RAG cuts token consumption by 23% and helps on the hardest queries.

What's Actually in a Litre of Diesel

Brent crude hit $119.50, then crashed to $88, then bounced to $90, all in 48 hours. Your fuel bill moved 5p. The anatomy of a diesel price, why wars move it less than you'd think, and why the Chancellor is still a bigger threat than Iran.

It's Just Weights

A tiny neural network reproduces the Cortical Labs CL1 DOOM demo in your browser. 132 parameters vs 200,000 biological neurons.

goformersearch: Vector Search in Pure Go

Brute-force and HNSW nearest-neighbour search with zero dependencies. When to use which, backed by data.

goformer: BERT Inference in Pure Go

Loading HuggingFace models without leaving Go. No CGO, no ONNX, no Python.

Bitwise Cloud: Hosted Search for Embedded Documentation

Taking the bitwise-mcp local search tool and turning it into a hosted multi-tenant platform with a web app, REST API, and plugin marketplace.

CO2 Is the Wrong Number: Greenhouse Gas Equivalents for Road Freight

Fleet dashboards report CO2. Government targets regulate CO2. But CO2 is only one greenhouse gas, and for some powertrain choices, it's not the one that matters most.

Every Euro VI Truck Has a Chemical Plant Bolted Underneath

Four catalysts, three chemical reactions, a 600-degree soot incinerator, and 44 sensors. The chemistry and thermodynamics of modern diesel aftertreatment, explained at the molecular level.

The Physics and Economics of Moving 44 Tonnes at 56mph

Weight limits, speed limits, fuel consumption at scale, and why trucks do the things that annoy you on the motorway.

The Hydrogen Truck Problem Isn't the Truck

Hydrogen fuel cell trucks exist. They work. The problem is everything around them: production efficiency, infrastructure economics, and a competitor called 'the grid' that's already everywhere.

50kWh In, 33kWh Out: On-Vehicle Hydrogen Generation

The energy chain behind HHO systems, worked from a real product's published specifications. Why on-vehicle hydrogen generation for diesel trucks is a thermodynamic impossibility, not an engineering challenge.

Playing DOOM in OpenSCAD at 10-20 FPS

DOOM in OpenSCAD at 10-20 FPS - real-time 3D rendering in a parametric CAD tool using the Manifold geometry kernel

Zookeeper: Pokédex for Real Zoo Animals

Pokédex for real zoo animals - AI-powered collection game that turns zoo visits into adventures with Gemini vision identification

Phaestus: Compiler for Physical Products

Compiler for physical products - natural language to manufacturable PCB, enclosure, and firmware using constrained AI generation

TOKN: Token-Optimised KiCad Notation

Token-Optimised KiCad Notation - 92% token reduction format for LLM circuit generation, with benchmark suite and model comparison

KiDoom: Running DOOM on PCB Traces

Running DOOM on PCB traces - vector-based rendering of the classic 1993 shooter using KiCad's PCB editor at 10-25 FPS

ScopeDoom: DOOM on an Oscilloscope via Sound Card

DOOM on an oscilloscope via sound card - using a MacBook's headphone jack as a dual-channel DAC to drive vector graphics at 4-8 Hz

CircuitSnips.com: Building a Thingiverse for KiCad

Thingiverse, but for KiCad subcircuits.

bitwise-mcp: MCP Server for Embedded Documentation

MCP documentation server for embedded developers. Ingests PDF RMs, extracts register definitions, and provides fast semantic search with minimal token usage.

Engine Simulator: High-Fidelity Diesel Engine Thermodynamics

Engine Simulator, using physics calculations, currently set up for diesel 6cyl

Claude Code Comm Bot: VS Code Extension for Discord Integration

Discord Communication bot for Claude Code, VS Code integration

Fuel Injector Characterizer: Peak & Hold Validation System

One and Done, Vibe Coded system for validating fuel injector performance. Arduino and Webserial.

KiCad Netlist Tool

Tool for reducing LLM token count from Kicad SCH files, to allow for automated documentation

MQTT Duration Trigger: Smart Button Press Detection

Home Assistant integration for MQTT-based duration triggers

Supplement Stack Saver: Budget-Friendly Alternatives

The supplement industry thrives on branding and marketing, often charging premium prices for common vitamins and minerals. Supplement Stack Saver cuts through the marketing to reveal a simple truth...

TheDuck.chat

AI-powered chat application. Developed for the T3 Cloneathon. DB currently inactive, demos available in repo.

HIT Impact Android App: Head Impact Monitoring

Kotlin Android app project for HIT Impact V1 traumatic brain injury sensors