In a recent interview at KubeCon and CloudNativeCon in London, Matt Butcher, CEO and Co-Founder of Fermyon, and Jon Alexander, SVP of Products at Akamai, discussed the impressive growth of WebAssembly (Wasm), its expanding adoption beyond browsers, and how their partnership is revolutionizing edge computing.

The Rise of WebAssembly Beyond Browsers

Once a niche browser technology, WebAssembly has evolved into a powerful foundation for next-generation computing—from serverless to edge-native, and now to AI. As Butcher explained: “WebAssembly started as a web browser technology and was very successful in that role, but it was so well constructed that it began to find uses outside of the browser.”

Its real promise lies in its design—a platform-neutral, binary execution format capable of running securely and efficiently across diverse environments, from browsers to servers to IoT devices. “WebAssembly is clearly a powerful technology that enjoys industry support—from the most emerging startups to the long-established stalwarts,” said Butcher.

The technology has seen adoption across various domains:

• Serverless runtimes (Fermyon’s focus)
• Embedded media applications (BBC’s media player)
• IoT implementations
• And now, edge computing

A significant milestone in WebAssembly’s evolution was Oracle‘s recent announcement of Java support. “Java was kind of the last major language to move over to WebAssembly,” noted Butcher.

Solving Performance Challenges in Serverless Computing

Traditional serverless platforms like AWS Lambda or Azure Functions have performance limitations, with 200-500 milliseconds of cold start time before code execution begins. This makes them unsuitable for high-performance edge use cases that require near-instantaneous response times.

Fermyon set out to “reinvent serverless” by building a highly performant runtime platform that maintains the feature richness of established serverless offerings. The partnership with Akamai combines Fermyon’s fast compute engine with Akamai’s global distributed network, delivering what Butcher describes as “mind-boggling performance” for edge applications.

Edge-Native Applications

Alexander explains how Akamai, traditionally an edge company focused on content delivery and security, is now expanding into edge computing: “When we talk about edge native, we’re referring to applications that benefit from running at the edge—being distributed, having low latency, running in specified territories. It’s very geography-centric, highly distributed, and performance-sensitive.”

The Akamai-Fermyon partnership allows for complete applications to run at the edge, eliminating the need for upstream data center components in many cases. This represents a significant shift in application architecture.

Diverse Use Cases Across Industries

The partnership has unlocked numerous use cases that were previously impractical to implement at the edge:

• AI inference
• Virtual reality applications
• E-commerce personalization
• Content watermarking
• Piracy detection
• Real-time data leak prevention

Alexander mentions they’ve accumulated “a list of 260 use cases” across virtually every industry. “The fact that it’s WebAssembly under the hood is less relevant than the fact that we’re allowing customers to run powerful code at the edge. That’s what really excites Akamai.”

Democratizing AI Through Edge Computing

AI applications are often constrained by their expensive, complicated implementation requirements. The Akamai-Fermyon partnership aims to address this by working toward “inference as a service” rather than the current model of “renting infrastructure, renting GPUs, deploying for peak and having a lot of resources sat idle.”

Perhaps one of the most compelling outcomes of this partnership is the enablement of AI at the edge. “AI inference—that’s a perfect fit,” Alexander said, referring to the compute-heavy, latency-sensitive workloads required for real-time, interactive AI applications. Akamai aims to simplify this by presenting “inference as a service,” which abstracts the complexity of managing GPUs and infrastructure from developers.

Alexander explains their approach: “A serverless infrastructure helps with that—it scales up and down—but it’s also important to have the infrastructure in the right place. Many of the AI applications we’re seeing are performance-sensitive. They’re latency-sensitive, interactive, and need to run in locations close to the end user, where the data also resides.”

The Future of WebAssembly

WebAssembly appears to be entering a period of stability after rapid evolution. With the WebAssembly System Interface (WASI) specification nearly complete and major language communities adopting the technology, Butcher believes WebAssembly has “hit that kind of stable point where WebAssembly is going to be a known quantity.”

The true innovation is now happening in how WebAssembly is being deployed across computing environments. “What we’re seeing is a blending of server edge, CDN edge, and IoT embedded devices into a continuum of computing,” says Butcher. “Because WebAssembly is a platform-neutral binary format, we can run it across all these different hardware profiles and operating systems.”

This openness and portability are key selling points for customers. As Alexander points out: “When they decide to run on Akamai, they’re also running in other environments—on bare metal, and perhaps writing client-side code as well. Having the ability to run across multiple locations without being tied to a single vendor is extremely important to them.”

What’s in the pipeline

The partnership between Akamai and Fermyon represents a significant advancement in edge computing capabilities. By combining Fermyon’s WebAssembly expertise with Akamai’s global network, they’ve created what Butcher calls “the fastest serverless platform in the world—running on the fastest, most distributed network in the world.”

As WebAssembly continues to mature and gain industry-wide adoption, we can expect to see increasingly sophisticated applications running at the edge, transforming how developers architect and deploy their solutions.