SSVM VS e2core

WasmEdge GitHub address: https://github.com/WasmEdge/WasmEdge.

I see that they recently changed the intro of WasmEdge on Github [1] to " WasmEdge is the easiest and fastest way to run LLMs on your own devices. "

Since it's a wasm runtime capable of many things I find bizarre that they now start describing it with a ultra-specific use case

- [1] https://github.com/WasmEdge/WasmEdge

Mmm…

The wasm-nn that this relies on (https://github.com/WebAssembly/wasi-nn) is a proposal that relies of arbitrary plugin backends sending arbitrarily chunks to some vendor implementation. The api is literally like set input, compute, set output.

…and that is totally non portable.

The reason this works, is because it’s relying on the abstraction already implemented in llama.cpp that allows it to take a gguf model and map it to multiple hardware targets,which you can see has been lifted here: https://github.com/WasmEdge/WasmEdge/tree/master/plugins/was...

So..

> Developers can refer to this project to write their machine learning application in a high-level language using the bindings, compile it to WebAssembly, and run it with a WebAssembly runtime that supports the wasi-nn proposal, such as WasmEdge.

Is total rubbish; no, you can’t.

This isn’t portable.

It’s not sandboxed.

If you have a wasm binary you might be able to run it if the version of the runtime you’re using happens to implement the specific ggml backend you need, which it probably doesn’t… because there’s literally no requirement for it to do so.

There’s a lot of “so portable” talk in this article which really seems misplaced.

WasmEdge isn't there yet: https://github.com/WasmEdge/WasmEdge/issues/1122#issuecommen...

Example 1: Analyze the content and risks of each commit in the PR. Then make a summary. https://github.com/WasmEdge/WasmEdge/pull/2394#issuecomment-...

There has also been a few "cloud-native" runtimes based on WASM, like WasmEdge but there's a few others (can't remember their names!)...

> If one writes Go or Rust, there are much better ways to run them than targeting WASM

wasm has its place, especially for contained workloads that can be wrapped in its strict capability boundaries (think, file-encoding jobs that shouldn't access anything else but said files: https://news.ycombinator.com/item?id=29112713).

> Containers are still the defacto standard.

wasmedge [0], atmo [1], krustlet [2], blueboat [3] and numerous other projects are turning up the heat [4]!

[0] https://github.com/WasmEdge/WasmEdge

[1] https://github.com/suborbital/atmo

[2] https://github.com/krustlet/krustlet

[3] https://github.com/losfair/blueboat

[4] https://news.ycombinator.com/item?id=30155295

One of the first things we've used it for internally is to run one-off isolated tests on WebAssembly modules instead of feeding them through a production Atmo[0] instance. It basically serves as a dumb pipe for feeding data in and out of a Wasm module.

0: https://github.com/suborbital/atmo

Excited to see Atmo on there 🙂 https://github.com/suborbital/atmo

My open source focus for this year is building Atmo, and there is one aspect of the process that I would like to highlight. Since early 2020 I knew roughly what I wanted to build. The specifics of that thing changed over time, but the core idea of a server-side WebAssembly platform was consistent all throughout the year. I didn't write a single line of code for Atmo until late October, even though that was what I wanted to build the entire time. I want to talk about why.

I am also open to any and all contributions from the community. I am more than happy to meet with anyone interested in working alongside me to build these capabilities so that I can help get you started developing Atmo, Vektor, Grav, Hive, and Subo. Developers with no experience working with WebAssembly, distributed systems, web services, or Go are encouraged to join and I will do whatever I can to help you learn what's needed to contribute. Open Source is not just about developing in the open, it's also about helping others learn.

With SUFA systems, multiple ASGs are created, each designated as a capability group. Each capability group is given access to the resources required for the associated function namespace to operate (such as the datastore or secrets), and can then scale independently of one another. Since the application's functions are decoupled entirely from one another, it's possible for some functions to run on the host that receives the request, and functions from particular namespaces to be meshed into other capability groups. A SUFA framework such as Atmo is responsible for handling the meshed communication, completely absorbing the complexity.

The SUFA pattern was designed in concert with Atmo, which is an all-in-one framework upon which SUFA systems can be built. Atmo uses a file known as a 'Directive' to describe all aspects of your application, including how to chain functions to handle requests. You can write your functions using several languages to be run atop Atmo, as it is built to use WebAssembly modules as the unit of compute. Atmo will automatically scale out to handle your application load, and includes all sorts of tooling and built-in best practices to ensure you're getting the best performance and security without needing to write a single line of boilerplate ever again.