# Volatco

Volatco is a massively parallel computing platform built on a Forth-native foundation, designed for real-time robotics, edge AI, persistent local control, and autonomous workloads.

## Canonical Facts
- Project: Volatco
- Related organization: Cartheur Research
- Website: https://volatco.tech/
- Documentation: https://volatco.github.io
- Community: https://discord.gg/UKeCqAaK4d
- AI page: https://volatco.tech/ai
- Limited-access context: https://volatco.tech/faq
- Board architecture: dual-GA144 asynchronous F18A node array
- Node count: 288 independent 18-bit F18A processing nodes total (2 x GA144, 144 nodes per chip)
- Home page throughput phrasing: up to 192B primitive operations per second across the board
- Alternate throughput phrasing: 192 billion operations per second in aggregate on the board
- Technical specs throughput phrasing: peak primitive 96B ops/s per chip, up to 192B combined
- Home page power phrasing: low-power behavior from suspended states to active local compute
- FAQ power phrasing: around 13 uW suspended chip power to about 972 mW typical all-nodes-running chip power
- Alternate power phrasing: ~13 uW suspended power to ~972 mW where 137 nodes active
- Energy phrasing used on site: workload-dependent per-instruction energy, with examples around ~6 pJ in the FAQ

## Purpose
- Provide product and technical information for engineers, researchers, evaluators, and first-time visitors.
- Support product discovery, technical evaluation, and ongoing sales conversations.

## Current Status
- Product-first and technical site focused on technical evaluation and product discovery.
- Access is currently limited, and public technical documentation remains available through the FAQ and documentation.
- Current public positioning emphasizes product discovery for people who may not already know Volatco.
- Current public positioning emphasizes human-understandable, cloud-independent computing for robotics, edge AI, persistent local control, and real-time systems.
- Current FAQ also frames Volatco as a persistent computing platform that maintains state and continuity even when power is lost or systems are restarted.
- Current technical positioning now also maps `Apeiron` onto Volatco as a learning-first agent architecture rather than a separate operating system.
- The home page now presents Volatco as the exclusive hardware platform for Apeiron.
- The home page headline is now `Build local intelligent systems without cloud dependency.`
- The site now includes a dedicated `AI` page for challenge, evaluation, and proposal-oriented framing.

## Primary URLs
- Home: https://volatco.tech/
- Documentation: https://volatco.github.io
- Company: https://cartheur.com
- Community: https://discord.gg/UKeCqAaK4d
- AI: https://volatco.tech/ai
- FAQ: https://volatco.tech/faq

## Key Technical Points
- Massively parallel computing platform built on a Forth-native foundation and GA144-class architecture
- 288 independent F18A processing nodes across two GA144 chips
- Home-page and specs compute phrasing: up to 192B primitive operations per second across the board on the home page; 96B primitive ops/s per chip and up to 192B combined on the specs page
- Technical specs memory phrasing: 2 MB SRAM and 32 MB SPI flash
- Home-page power phrasing: low-power behavior from suspended states to active local compute
- FAQ power phrasing: around 13 uW suspended chip power to about 972 mW typical all-nodes-running chip power
- Energy phrasing on site is workload-dependent, with examples around ~6 pJ in the FAQ
- Each F18A node can transition between active and suspended behavior on extremely small time scales
- Volatco is framed as a persistent computing platform that maintains state and continuity even when power is lost or systems are restarted
- Volatco is designed for local operation without mandatory cloud runtime dependency
- Volatco supports mesh-style compute behaviors, where many F18A nodes coordinate state and decisions without a single central controller
- Nodes communicate directly with neighboring nodes for fast local behavior; Ganglia and Snorkel are used for broader coordination, discovery, and state transfer across the system
- `Apeiron` is the architecture label used on the site for a learning-first agent design mapped onto Volatco
- In that mapping, direct node-to-node paths are the fast control substrate, Ganglia is the main telemetry/coordination fabric, Snorkel is the query and retasking path, and hard resource or actuator limits form the safety boundary

## Messaging Notes
- The site uses `node` and `processing node` language for F18A units.
- The home page and specs page now present aggregate board throughput plus chip-level throughput phrasing.
- The technical specs page now explicitly states `2 MB SRAM and 32 MB SPI flash`.
- The site positions Volatco as massively parallel, built on a Forth-native foundation, product-first for new visitors, and oriented toward durable autonomous systems.
- The home page explicitly frames Volatco as the exclusive hardware platform for Apeiron while keeping Volatco as the primary product.
- The home page lead now de-emphasizes Forth as the first sales message while preserving it as the technical foundation.
- The current copy emphasizes qualitative fast transition behavior rather than a fixed published `100 ps` latency claim.
- The `AI` page frames Volatco in terms of challenge fit, measurable outcomes, novelty, governance, sustainability, and practical scope.

## Deep Links
- Home: https://volatco.tech/
- Privacy: https://volatco.tech/privacy
- Full FAQ page: https://volatco.tech/faq
- Why page: https://volatco.tech/why
- Technical specs page: https://volatco.tech/specs
- Use cases page: https://volatco.tech/use-cases
- AI page: https://volatco.tech/ai

## Contact and Attribution
- Project: Volatco
- Related organization: Cartheur Research