Open Source Summit NA 2025 | Sam Gammon, Elide

Clips
News
More from Open Source Summit NA 2025

Sam Gammon

Co-Founder & CEO

Code, community and the future: 13 takeaways from Open Source Summit NA

Beyond uniting developers under shared ideologies, goals and pain points, Open Source Summit North America ignited conversations at the intersection of code, community and change.TheCUBE goes live with insights at Open Source Summit NA

pause_circle_outline Allied aims to address the imbalance in the JavaScript ecosystem for other languages.

play_circle_outline Revolutionizing Performance: How Modern Runtime Enhancements Boost Speed and Interoperability Across Multiple Programming Languages

play_circle_outline Two cost-saving mechanisms: lowering developer salaries and reducing cloud compute expenses.

Info
Transcript

Sam Gammon, Elide

Sam Gammon

Co-Founder & CEO

In this interview from the Open Source Summit North America 2025, Sam Gammon, founder and developer behind Elide, joins theCUBE Research’s Paul Nashawaty to discuss why the future of software performance lies in rethinking the runtime. Built on Oracle’s GraalVM, Elide offers a radical new approach to language interoperability, compiler speed and system performance – delivering up to 75x gains over Node.js while remaining fully open source.

Gammon explains how Elide bridges the gap between traditionally siloed languages like Java, Python and JavaScrip... Read more

explore Keep Exploring

What is Allied and what motivated its creation? add

What are the performance comparisons of Graal VM with other programming languages like Python, Node.js, and Ruby, and how does Graal VM facilitate interoperability among different languages? add

What are the advantages of using React developers in a software development team compared to Java and Python developers? add

bolt Powered by CUBE AI

Sam Gammon, Elide

search

Paul Nashawaty

>> theCUBE research shows that new run times can deliver up to 75 times better performance using modern platforms. Welcome to this episode of the App Dev Angle, coming to you live from the Open Source Summit North America 2025. My name is Paul Nashawaty, covering the app dev space, and all things open source. I'm joined from Sam, from Allied.

Sam Gammon

>> Yes.

Paul Nashawaty

>> Sam, how you doing?

Sam Gammon

>> Hi, how's it going?

Paul Nashawaty

>> Great, great. Good take. Great to have you here.

Sam Gammon

>> Thank you for having me.

Paul Nashawaty

>> Want to tell the audience a little bit about yourself and about Allied?

Sam Gammon

>> Sure. So my name's Sam. I'm a software developer. We made Allied, because we were tired of seeing the JavaScript universe expand to the detriment of all others, and we feel like that expansion has really left a dearth of activity in say, Python or Java, these sort of back-end spheres. So we've made a runtime called Allied, which is basically Node, but from a Java first perspective.

Paul Nashawaty

>> That's awesome. I mean, when we look at this, we look at runtime platforms that are looking to redefine performance, cost efficiencies, and developer flexibility across multiple programming language, right? These polyglot deployments are really looking for new ways of doing this, especially with these heritage environments. So I understand that you're built on Oracle Graal VM, right? This is a solution that promises dramatic speed improvements, so up to 70 times faster than traditional runtimes, right? We were trying to say that upfront, right? But, while remaining open source and drop in-ready for existing ecosystems, there's a lot here to think about, but these are some lofty claims, Sam, right?

Sam Gammon

>> Yes, yes, they are.

Paul Nashawaty

>> You think about that. So why don't we tell the audience a little bit about that?

Sam Gammon

>> Sure. So lofty claims require strong evidence, right, and we're independently benchmarked. So the numbers that I'm going to give you here are either from Oracle, or from a group called TechEmpower, which independently benchmark servers. So we're three times faster than Stock Python, we run Python. We're 75 times faster than Node.js at running JavaScript. We're 24 times faster at running Ruby. And all of this is based on the technology that Oracle has created with Graal VM. These language engines are brand new, they're really well written, they're modern, and they interoperate. So in our paradigm, you can actually use multiple programming languages together. They can import from each other, they can call each other. There's no inter-op boundary where there's overhead added, right? There's no serialization, there's no process boundary, and I'm getting very technical here, but the point is that it's already the case that many Java Code bases or Python code bases also have node in them.

Paul Nashawaty

>> Right.

Sam Gammon

>> They have a bunch of node modules react, that they want to add for their UI. This lets you put all of that into one system, one tool chain, and use it together.

Paul Nashawaty

>> So when we think about this interoperability across these different polyglot languages, the use case has benefits from these multi-languages' capabilities can really be exponential, right?

Sam Gammon

>> Totally. Exponential.

Paul Nashawaty

>> Can you talk about some of these use cases?

Sam Gammon

>> Sure. So one big use case for PolyGlot, and PolyGlot isn't necessarily Elide's reason to exist. It's just a natural benefit that out of this new model, right? But one big benefit of PolyGlot is we can install packages from Maven, from PYPI, from NPM, and so, all of a sudden, your developers have much more software that they can pull off the shelf, to do something without building themselves, right? It nearly triples the ecosystem that you can pull from very easily, to use in your app, right?

Paul Nashawaty

>> Okay. So I open up the session with 75 times more performance, right? That was kind of a little bit of a struggle, but I got it out there. But when I think about that, this is across various languages. Can you explain that number, and how you got there?

Sam Gammon

>> Sure. So some of our benchmarks, like the 3X Python stat, the 24X Ruby stat, those are the geometric mean of a language suite of benchmarks that test everything from function calling, to arrays, to iteration. Then, on the server side, where we have JavaScript like numbers like the 75X, that's a server benchmark, so that's requests per second, right? And Elide, as an example, can sustain 800,000-plus requests per second on Linux, where Next.js on Node can only do 3,000, right? And so, this is a night and day difference, hundreds of thousands of more, at RPS.

Paul Nashawaty

>> And how do people know about this?

Sam Gammon

>> Well, we're open source. We're on GitHub.

Paul Nashawaty

>> Okay.

Sam Gammon

>> We've just launched beta about a month and a half ago, and we've seen a great response so far. Generally, when developers find us, they star us, they follow us, they want to know when this is going to be ready. And I think stable for us is probably about three months away.

Paul Nashawaty

>> Okay. So let's talk about that, because when we think about that, we want to make sure that we put this out there, we get people to understand what it can do. You're kind of in the, I'll call it an alpha stage, at this point?

Sam Gammon

>> Beta for sure.

Paul Nashawaty

>> Beta for sure? Okay, okay. That's fair. That's a good way to put it.

Sam Gammon

>> Bugs are unexpected, yeah, yeah.

Paul Nashawaty

>> Of course, but that's just the way, the nature of the business.

Sam Gammon

>> Yeah.

Paul Nashawaty

>> But as a drop-in replacement for Node and Node.js, what challenges did you face with the compatibilities between the Node ecosystem, and the partners that kind of ensure that parity, between the libraries and the tools?

Sam Gammon

>> Sure. Great question, actually. So we have the same fight that new alternative Node runtimes have, which is, you think of Deno and Bun, all of these companies kind of have to recreate the Node API?

Paul Nashawaty

>> Yup.

Sam Gammon

>> We have to do that as well. But one big benefit of Elide is, we're built on the JDK. So a good example would be reading a file. When it comes to reading a file, if you want to do that in native code, you have to do that on six operating systems, and efficiently, and it's very hard to do. So they've got their work cut out for them. We just read a file in the JDK, which is really fast, really well-tested, works on every operating system. So yeah, our fight is a lot easier than the other ones going on right now.

Paul Nashawaty

>> So why haven't people done this before?

Sam Gammon

>> That's a great question. I think it's because Node was written by Ryan Dahl, who was a particular fan of C++ and JavaScript, and that's sort of what it took for somebody to come along, and take V8, and put it on the server with the APIs needed in JavaScript. We have that same bent. We love all these languages, but we're Java Kotlin first, right? And we believe that the JVM, and Java and Kotlin give a richer, more powerful substrate to build a runtime in. Here's an example. Java's memory safe. And that means that most of Elide is also memory safe, right?

Paul Nashawaty

>> Okay.

Sam Gammon

>> Just by natural default. And so, we think about a world where 85% of vulnerabilities and exploits today, somewhere along the line, come from memory safety. Right? I mean, we could eliminate all of those in one go, by switching to a better runtime.

Paul Nashawaty

>> Yeah, that's interesting. No, I think that makes a lot of sense. And when you think about alternate runtimes, it's customer choice. People are figuring out what they want to do, and how they want to get there. They also have to look at these barriers that they need to overcome, right? Part of it is thinking about it from their perspective of, we just recently had our App Dev Summit, and we did a number of bits of research around day zero, day one, day two, and build release and operations. One of the things we were thinking about was around the compilers, and what's happening in the compilers. And when we think about it, you make a claim of 20 times faster with the compiler here. Right? And that is interesting, because I think it also impacts some of the slowdowns you see in the CI/CD pipeline, right?

Sam Gammon

>> That's right.

Paul Nashawaty

>> So, if we look at it from that lens, one, validate the claim, right?

Sam Gammon

>> Yeah.

Paul Nashawaty

>> Let's talk about that.

Sam Gammon

>> Yeah.

Paul Nashawaty

>> And two, let's talk about the CI/CD pipeline, and how you can get the release out faster.

Sam Gammon

>> Sure. Sure. So our Java compiler and our Kotlin compiler are just embedded versions of the standard Java and Kotlin compiler. And what we've done is, we've made these native, rather than running on JVM, and as a result, there's no warmup time. So it's an identical compiler, but it doesn't take time to warm up anymore, right?

Paul Nashawaty

>> Okay.

Sam Gammon

>> So that's where we get all of our speed gains, and that's how we can hit exact same inputs, exact same outputs as the standard compilers, because it is that compiler, just treated differently, right?

Paul Nashawaty

>> I see. Okay.

Sam Gammon

>> And so, that's one answer to your question. Another answer to your question is safety and security, and all of those guarantees still apply in the compiler phase. Now, I should say, the 20X stat is the fastest case, right?

Paul Nashawaty

>> Okay.

Sam Gammon

>> I think users are more likely to see a 10 to 15X gain, but that's still incredible.

Paul Nashawaty

>> Well, yeah, of course it is.

Sam Gammon

>> Yeah.

Paul Nashawaty

>> I mean, the gain is, whether it's 10 or 20, I mean, you're talking about a significant gain free, right?

Sam Gammon

>> Yeah, yeah, absolutely.

Paul Nashawaty

>> So, from a tools and integration perspective, developers are using their bespoke solutions. They have certain tool sets that they're using, the tech stack they're using. How does this platform have minimal friction, especially when you're looking at existing Node.js or JVM-based environments?

Sam Gammon

>> Sure. Great question. So we have plugins for Gradle and Maven. We prefer to meet users where they are, which means, we're going to integrate with your build system. We're going to give you APIs that you're familiar with, before we design our own new APIs, or break you, right? Our goal is really to just meet you where you are, and do a better job than your current runtime is doing. So we're fully compatible with Java, fully compatible with Python. Node is the only one where we have this longer fight to get to parity, right?

Paul Nashawaty

>> Gotcha.

Sam Gammon

>> Yeah.

Paul Nashawaty

>> Okay. I mean, it makes sense. So it's an evolution. It's going to get there over time, but your gains that you're seeing is definitely worth the investment.

Sam Gammon

>> Oh, users can benchmark themselves.

Paul Nashawaty

>> Sure.

Sam Gammon

>> I mean, all we do is we compile the same code with Java C and Elide, and it's Elide Java C. So you just put Elide in front of it, and see it different.

Paul Nashawaty

>> Yeah, absolutely, absolutely. So all right, let's pivot the conversation. We talked about it a lot on the practitioner side, right? We talked a lot about the developers, and how the developers are getting savings. Let's talk about the business decisions that are being made here, right? When we think about this new investment, right? How do we look at the platform, how it reduces software development costs, and cloud infrastructure costs? What types of ROI are you seeing with this?

Sam Gammon

>> So ROI is hard to measure on this, and I am not quite sure what my answer would be there yet, but I'll tell you the two mechanisms by which we save costs, or save costs for our users. Number one is, Java developers, Python developers, expensive.

Paul Nashawaty

>> Yeah.

Sam Gammon

>> React developers, plentiful and cheap. So how do you get a team of react developers and back-end developers to work together? This is a runtime that allows you to interoperate, that both their code could talk to each other. It's seamless. You don't have to throw it away, you don't have to rehire people, right? There's a cultural barrier that we're knocking down here. By doing that, we lower the cost to develop software, because those salaries can be lower for react and front-end developers. So that's one place.

Paul Nashawaty

>> Gotcha, okay, okay.

Sam Gammon

>> The other place is by running faster, we can run for less time. So that cuts your AWS compute bill immediately, right?

Paul Nashawaty

>> Yeah, well, I see that we can absolutely talk about this economic validation here, because I think it's a big factor in switching platforms, moving to a new tech stack, and such. I think that's an important factor to look at.

Sam Gammon

>> Absolutely.

Paul Nashawaty

>> But we will also want to talk about those tangible business values, right? Like user experience, you were talking about that, you have the collaboration between teams, faster releases, because you have that collaboration.

Sam Gammon

>> Faster builds.

Paul Nashawaty

>> Faster builds. We talked about that across the CI/CD pipeline.

Sam Gammon

>> Right.

Paul Nashawaty

>> But also revenue impact, right? Because I mean, you can get this code out the door faster, but anything else that we're missing, other than, I mean, the performance acceleration alone is incredible.

Sam Gammon

>> Well, one that's hard to quantify is, what would a breach cost you, that never happened, right? In a memory-safe runtime, you can be totally sure that there's not going to be some bug where they can escape the VM, and steal all of your company's data, and have a cataclysmic breach. That's an impossibility with us, right? And so, closing those gaps for enterprise is what we see as innovation.

Paul Nashawaty

>> Gotcha, gotcha.

Sam Gammon

>> Yeah.

Paul Nashawaty

>> All right, so I have a business question here. When I think about the MIT license here that kind of goes with this, right, how do you balance the openness between sustainability and the commercial side, of making money and support opportunities?

Sam Gammon

>> So our answer is open source first, because I think what we build has to be useful for the world. With open source feedback, and contributors and attention, we can make sure it is useful for the world. We want to be oriented towards everybody's needs here. On the commercial side, we have plans where we're going to be providing premium tool chains, in JVM and Android. So those are things that companies can purchase to accelerate their build, that's separate and different from our runtime. So the runtime's always going to be open source, always free, right?

Paul Nashawaty

>> Very cool.

Sam Gammon

>> Now, those components that we're intending to charge for, in all cases, we charge less than the cost that you would incur, for not using our product, right?

Paul Nashawaty

>> Right.

Sam Gammon

>> So they're almost always structured to be a no-brainer.

Paul Nashawaty

>> All right, so then, I'll ask the question. Say you're talking to CTOs here at the audience, right? And we're at open source, and they're evaluating these new solutions and how to get there. And you're looking at whether to migrate an existing workload to your platform, or migrate success stories around this. Do you have some success stories that you can share?

Sam Gammon

>> Not yet, we're very early. But here's what I would say, is that what we've seen work with some private design partners is dropping in Elide as a compiler replacement first, because brownfield projects already use these compilers. They can experience an immediate gain. There's no difference in the output that they receive. So that's just a freebie, right?

Paul Nashawaty

>> Gotcha.

Sam Gammon

>> That's where the brownfield projects come in. Greenfield, they have a whole new way to develop software, ship software, that's much cheaper than before.

Paul Nashawaty

>> Gotcha, gotcha. So we're sitting here a year from now, the next Open Source Summit, right? Where do you see this all going, and where's it taking off?

Sam Gammon

>> I think AI is here to stay, in a lot of ways.

Paul Nashawaty

>> Sure.

Sam Gammon

>> I think AI is going to be local and sovereign, when we get there. I think businesses are going to have their own I in-house. They're not going to be paying Anthropic or OpenAI for bazillions of dollars, just to run their business. And I think, to get there, we need to evolve on the runtime side. So Elide, for instance, can run local AI very easily. So we've made that super easy, so that your developers can take your private data, keep it sovereign on-site, but have the whole chat interface that you want, right? So yeah, I think in a year, we're going to be back here. We're going to have a booth. It's going to be the nicest booth, I hope.

Paul Nashawaty

>> That's great, that's great. Well, look, I mean, I think, congratulations on the successes you've had so far.

Sam Gammon

>> Thank you.

Paul Nashawaty

>> This is awesome. I really think you're going to give the industry a run for their money, right?

Sam Gammon

>> Oh, thanks.

Paul Nashawaty

>> I think it's a great, great thing. I mean, performance is incredible. And I want to thank you for your time and your insights, it's been great.

Sam Gammon

>> Thank you for having me.

Paul Nashawaty

>> And thank you for joining me today. I'm Paul Nashawaty, already coming to you live from the show floor at Open Source Summit North America 2025. Thank you to the audience for watching theCUBE, the leading source for tech news.