What is the tile platform?
Tiles are tiny, standardized hardware modules that snap together into real electronics. This page is the five-minute mental model — the handful of words you’ll see everywhere in these docs, and how they relate.
Tiles
A tile is a self-contained microelectronic module in a standard 4×4 mm package, with ten pads on a 0.8 mm pitch. Each tile packs a complete subsystem — a sensor, a motor driver, a power regulator, a radio — behind a common electrical and software interface. You wire tiles together instead of laying out chips, and the same tile carries a design from breadboard prototype all the way to mass production.
Browse them in the catalog: each tile lists its capabilities, pinout, and interfaces.
Cores run the show
Most tiles are peripherals — they sense or actuate, but they don’t think. A Core tile is the exception: it carries a microcontroller (an STM32) and runs your firmware. The Core talks to the other tiles over shared buses (I²C / I³C / SPI) and orchestrates the system.
Drivers and the PAL
Every tile ships with an open-source, platform-agnostic driver — C code that handles its registers, data formats, and protocol so you don’t have to. A thin Platform Abstraction Layer (PAL) bridges those drivers to whatever processor you’re on. Today the PAL targets STM32-based Cores; ports for ESP32, Arduino, and nRF are on the way.
- Driver — knows the tile (e.g.
tile_sense_bp). - PAL — knows the processor (how to do an I²C transfer on this host).
Two ways to build
You can build on tiles two ways, on the exact same hardware:
- Studio — a browser IDE where you compose a system visually, simulate its firmware, and build a binary. No toolchain, no C.
- The C SDK — write firmware directly in C against the
core_*APIs, with full control. Build and flash from your own machine.
Not sure which fits you? See Studio vs the C SDK.