Enterprise adoption of Rust language

Before switching to Rust, the engineering team at Ather used C and GO predominantly. The team is now looking at Rust as a stack to move towards for most of its use-cases. The engineering team uses Rust in some of Ather’s services on the edge device, which has shown significant improvement in resource consumption and memory saftey.

A micro service architecture on edge devices requires the stack to be interoperable across languages and devices. Rust has excellent support for this across stacks. Rust is also something that engineering teams can use for building on micro controllers and microprocessors.

The POCs Ather is evaluating in Rust is mainly in terms of performance benefits, maintainability, TAT and interoperability due to a larger community support and a plethora of crates serving different purposes.

The major concern to move to Rust is that it has a steep learning curve, which is a trade-off that the team is willing to make to enable a stable and performant ecosystem. Every ounce of resource available, matters on an edge device- CPU, memory, disk.

The first choice for performance is always C. But as the product is maturing and growing, there is more scope for edge computation coming in for different use-cases. As the scope grows, the need for quicker turnaround on POCs and features becomes as important as performance. This is where Rust looks to be the right choice for the team at Ather. Rust primarily offers equivalent performance benefits as C and also helps in quicker development time through its features like compile time memory safety check, borrow check, detailed errors, etc.

Rust generates relatively smaller binaries than GO, and has very low footprint. Rust’s ownership model has ensured memory safety; there have been no memory leakages till date on streaming applications.