firecracker VS Servo

This architecture leverages microVMs for rapid scaling and high-density workloads. But does it work for GPU? The answer is no. You can look at the old 2019 GitHub issue and the comments to it to get the bigger picture of why it is so.

Hi, I'm the CEO of the company that built this SDK.

We're a company called E2B [0]. We're building and open-source [1] secure environments for running untrusted AI-generated code and AI agents. We call these environments sandboxes and they are built on top of micro VM called Firecracker [2].

You can think of us as giving small cloud computers to LLMs.

We recently created a dedicated SDK for building custom code interpreters in Python or JS/TS. We saw this need after a lot of our users have been adding code execution capabilities to their AI apps with our core SDK [3]. These use cases were often centered around AI data analysis so code interpreter-like behavior made sense

The way our code interpret SDK works is by spawning an E2B sandbox with Jupyter Server. We then communicate with this Jupyter server through Jupyter Kernel messaging protocol [4].

We don't do any wrapping around LLM, any prompting, or any agent-like framework. We leave all of that on users. We're really just a boring code execution layer that sats at the bottom that we're building specifically for the future software that will be building another software. We work with any LLM. Here's how we added code interpreter to Claude [5].

Our long-term plan is to build an automated AWS for AI apps and agents.

Happy to answer any questions and hear feedback!

[0] https://e2b.dev/

[1] https://github.com/e2b-dev

[2] https://github.com/firecracker-microvm/firecracker

[3] https://e2b.dev/docs

[4] https://jupyter-client.readthedocs.io/en/latest/messaging.ht...

[5] https://github.com/e2b-dev/e2b-cookbook/blob/main/examples/c...

As far as I know, Fly uses Firecracker for their VMs. I've been following Firecracker for a while now (even using it in a project), and they don't support GPUs out of the box (and have no plan to support it [1]).

I'm curious to know how Fly figured their own GPU support with Firecracker. In the past they had some very detailed technical posts on how they achieved certain things, so I'm hoping we'll see one on their GPU support in the future!

[1]: https://github.com/firecracker-microvm/firecracker/issues/11...

I pass through a GPU and USB hub to a VM running on a machine in the garage. An optical video cable and network compatible USB extender brings the interface to a different room making it my primary “desktop” computer (and an outdated laptop as a backup device). Doesn’t get more silent and cool than this. Another VM on the garage machine gets a bunch of hard drives passed through to it.

That said, hardware passthrough/VFIO is likely out of the current realistic scope for this project. VM boot times can be optimized if you never look for hardware to initialize in the first place. Though they are still likely initializing a network interface of some sort.

“MicroVM” seems to be a term used when as much as possible is stripped from a VM, such as with https://github.com/firecracker-microvm/firecracker

According to their own FAQ it is indeed: https://github.com/firecracker-microvm/firecracker/blob/main...

What about microVMs like firecracker?

Dynamic memory management - Firecracker's RAM footprint starts low, but once a workload inside allocates RAM, Firecracker will never return it to the host system. After running several workloads inside, you end up with an idling VM that consumes 32 GB of RAM on the host, even though it doesn't need any of it.

Firecracker has a balloon device you can inflate (ie: acquire as much memory inside the VM as possible) and then deflate... returning the memory to the host.

https://github.com/firecracker-microvm/firecracker/blob/main...

Very few things in those companies are being written in Rust, and half of those projects chose Rust around ideological reasons rather than technical, with plenty of 'unsafe' thrown in for performance reasons

https://github.com/firecracker-microvm/firecracker/search?q=...

The fact that 'unsafe' even exists in Rust means it's no better than C with some macros.

Don't get me wrong, Rust has it's place, like all the other languages that came about for various reasons, but it's not going to gain wide adoption.

Future of programming consists of 2 languages - something like C that has a small instruction set for adopting to new hardware, and something that is very high level, higher than Python with LLM in the background. Everything in the middle is fodder.

https://github.com/firecracker-microvm/firecracker is the one that comes to mind, but most of these are internal.

> Is there any easy to use/hack HTML layouting engine where I could experiment with custom CSS attributes and bridge that gap? Would anything from Servo be suitable?

Servo could be used for this. You'd want to add support for parsing the CSS properties themselves to the style crate in https://github.com/servo/stylo and then the layout implementation to the layout2020 crate in https://github.com/servo/servo. You do effectively get a whole browser though.

I'm currently working on building a lighter weight / hackable layout engine based on a combination of https://github.com/servo/stylo (for css parsing and selector resolution), https://github.com/DioxusLabs/taffy (for box-level layout) and https://github.com/pop-os/cosmic-text (for flow/inline layout). I expect to have something decent in around 6 months

Neither of these setups currently have any support for pagination though.

Great to see some competition still alive in browser engine development. See also Servo (previously part of Mozilla) https://servo.org/ - that and Ladybird are still very underdeveloped compared to every day browsers.

It's a huge shame that there are no nightly builds of ladybird to try out but I assume that's because they just don't want the bug reports (if everything doesn't work it's pointless getting random bugs filed).

I haven't messed with it yet but from looking into it, this should absolutely work.

https://github.com/servo/servo/wiki/Building-on-ARM-desktop-...

don't know, there was a downtime in 2021 and 22 but since 2023, contributions look back to where it was before .. https://github.com/servo/servo/graphs/contributors

The world has moved on though to opinionated tools, and Rust isn't even the furthest in that direction (That would be Go). The equivalent of those two lines in Cargo.toml would be this example of a basic configuration from the jacoco-maven-plugin: https://www.jacoco.org/jacoco/trunk/doc/examples/build/pom.x... - That's 40 lines in the section to do the "defaults".

Yes, you could add a load of config for files to include/exclude from coverage and so on, but the idea that that's a norm is way more common in Java projects than other languages. Like here's some example Cargo.toml files from complicated Rust projects:

Servo: https://github.com/servo/servo/blob/main/Cargo.toml

rust-gdext: https://github.com/godot-rust/gdext/blob/master/godot-core/C...

ripgrep: https://github.com/BurntSushi/ripgrep/blob/master/Cargo.toml

socketio: https://github.com/1c3t3a/rust-socketio/blob/main/socketio/C...

1. Servo

It won't don't worry. There already are forks, for the worst case scenario. And Servo is on its way. Not yet ready, but it will be. Originally, from Mozilla kitchen.

To pain broad strokes, the layout phase (~= take the HTML, take the CSS, determine the position and size of boxes) is largely sequential in production browser engine today. Selector matching (~= what CSS applies to what element) is parallel in Firefox today, via the Stylo Rust crate originally developed in the research browser engine Servo. Servo can do parallel layout in some capacity (but doesn't implement everything), https://github.com/servo/servo/wiki/Servo-Layout-Engines-Rep... is an interesting and recent document on the matter.

Parallel layout is generally considered to be a complex engineering problem by domain experts.

https://hacks.mozilla.org/2017/08/inside-a-super-fast-css-en... is a really cool article that is related, that is a few years old but what it says is largely correct today.

[Article author/submitter here]

I can only tell you that it is not what this is about, inasmuch as I was at the talk and there was not a single mention of Firefox Reality or Wolvic in the talk.

Wolvic might use Servo – but I think if it did they would mention it, right?

The talk didn't and the word "Wolvic" does not occur anywhere on https://servo.org

So I am guessing not, no.

Igalia has -- or rather is because it's a co-op -- about 100 developers. They are not all working on the same thing.