AI as a service
Fast and safe AI inference in Node.js, written in Rust. Use pre-trained Tensorflow models to detect faces in photos.
Privacy-first and scalable data exchange service over the web. A Mozilla Open Labs project.
Machine learning in Rust could be 25x faster than Python. Detect patterns and visulaize your data.
As hardware miniaturization hits the quantum limit, software performance becomes the driver of growth. WebAssembly is a key technology that makes software more effcient, while preserving safety, portability, and software engineering best practices we built up in the past 25 years.Learn more Security Performance
Starter code templates
- A Substrate project template to create a new blockchain with the next-generation Ethereum virtual machine for executing smart contracts.
Get started with Rust functions in Node.js apps. Experiment with it without having to download any software or toolchain. Just open the repository in VSCode Codespaces.
Encryption and decryption are computationally intensive tasks for web apps. In this example, we will show you how to perform those tasks efficiently and safely in Rust and WebAssembly.
SSVM supports the WebAssembly System Interface (WASI) standard. It allows WebAssembly bytecode programs to access system resources such as the random seed, standard I/O, and the file system. Just make those system calls through the standard libraries in your Rust code. The ssvmup tool compiles them into WebAssembly to run in SSVM.
Built on the capability-based security model, the SSVM provides a Rust API to access AI hardware, such as inference chips. The Rust function calls are translated to hardware native operations and executed on the hardware. It boosts AI performance by 1000x and enables secure sharing of hardware among isolated applications.
SSVM provides a Rust API to safely call native programs. The WebAssembly function can pass data to and from the native program. The native program is typically a thin wrapper around a native library that is not yet supported in WebAssembly. The example is to call the tensorflow GPU native library from WebAssembly.