AltVMS: Move Over EVM

Maybe the EVM is a little like capitalism. So widely accepted, we struggle to envision a different future.

“It is easier to imagine an end to the world than an end to capitalism” - Frederic Jameson

As such, it wasn’t surprising to witness countless Alt-L1s pivoting to becoming Layer-2 during the height of L2 mania. Even Ripple launched an EVM layer - despite having a faithful army of XRP believers who never cared about EVM compatibility, to begin with. 

Fortunately, in the same way that philosophers propose alternative visions for the future, there are builders in the blockchain industry who refuse to accept the status quo simply because it’s convenient.

Frustration with the EVM has given rise to a vibrant landscape of AltVMS. During a recent space, the Subsquid team went live with such builders, developing across different alternatives. 

Speakers: Vitali (building Spark), Mikko (Founder Trading Strategy), Joe (Founder RLO and Contributor Near Dev Hub) and the Subsquid CEO Dmitry. 

Read on for a holistic overview of what AltVMs are, the flaws of EVM, and insights from the guests. You can also listen to the recording of the discussion here. 

A sensible start for any discussion around AltVMs is the one in which they are all contrasted with the EVM. 

What is the EVM? 

The EVM, short for Ethereum Virtual Machine, describes the execution environment on top of Ethereum. It’s where smart contracts are executed and deployed. The EVM works similarly to any Virtual Machine you could run on your PC, for example, to play a game from the 90s - emulating the tech from back then. It’s also complete, meaning it can execute any type of computation as long as it's provided with sufficient resources (time and memory). 

The introduction of fees (gas) ensures that inefficient code cannot run forever while preventing spam and other attacks. The downside is the more complex an operation is, the more expensive it gets. 

When a dev creates a smart contract they write it in a high-level programming language like Vyper or Solidity. Despite the name, smart contracts are anything but smart, they are more like complex if/then functions automating specific outcomes. 

The code is then compiled into bytecode understood by the EVM, which turns it into instructions and executes them. 

In 2014, the EVM enabled the first smart contract platform, or, as Dmitry called it, a “0 to 1 moment.” For some time, it was the one platform everyone built their projects on, leading to our status quo of EVM dominance. 

As the first VM after Bitcoin, Ethereum paved the way for composable Finance. 

“It was a milestone for all of us” - Vitalik 

That’s why all the VMs that aren’t EVM are called AltVMs now. But just because something is dominant doesn’t mean it’s great. 

Limitations of the EVM 

Mikko, one of the earliest developers to build with EVM, pointed out the obvious: EVM is ten years old tech, which was bad back then and is bad now. Everything else that came after has objectively been an improvement. 

A few of the reasons it’s bad are: 

• Security risks: With one VM so widely adopted, one bug easily affects multiple chains, and the countless hacks and breaches we’ve witnessed are a testament to that.
• Performance limitations: As also mentioned during the space discussing Parallel EVMs, the EVM doesn’t scale well because it processes transactions like the Aldi cashier: one after the other. This limits scalability and creates an environment where gas costs grow as volumes grow. 
• Solidity: requiring a custom programming language for a VM limits the potential developer community. On top of that, it’s considered complex and challenging to master while also having well-known flaws, such as reentrancy attacks and integer overflow, which have frequently been exploited. 

What’s more, especially for DeFi projects, whenever operating on the EVM, “you always chase TVL, requiring users to lock funds in an AMM model. A honeypot for Lazarus” (Vitali). 

With the rise of Layer 2s, there is also the question of how these will impact overall network security. What’s even more challenging is the fragmentation of UX and liquidity. Have you ever tried adding yet another L2 to your MetaMask and struggled? 

As a sensible person then looking for a place to build, you’d at least consider alternatives..right? 

So what are alternatives? 

There are countless AltVMS that are tailored toward specific needs, fulfilling requirements such as: 

• Language diversity: relying on more common programming languages 
• Higher scalability and speed
• Optimized Performance 
• Increase reliability and security features.

During the space, specific examples discussed were: NEAR, Solana, and Fuel. 

NEAR 

Joe from Near DevHub shared that NEAR chose WASM for the VM, focusing on performance and speed to provide a more universal runtime. An additional benefit is that it provides a more “web2” experience for builders and enables things such as yielded execution, where a contract makes calls and waits for a response instead of constantly re-calling.” 

Solana & the SVM 

Solana has been the most successful alternative L1, which is largely due to its innovative design, resilience, and its parallelization engine sealevel. Thanks to sealevel, transactions that are independent of each other are executed in parallel, increasing performance and efficiency. More interestingly, Solana can even be leveraged to help scale the EVM, as NEON EVM is demonstrating. 

Fuel 

Fuel is a purpose-built, custom virtual machine for executing smart contracts, learning from the EVM, and implementing improvements such as minimizing state, parallel transaction execution, and running on multiple threads and cores of CPUs than possible in other blockchains. Fuel also introduced a custom language combining the syntax of Rust and similar safety guarantees with elements from Solidity. 

These are just a handful of existing alternative VMs. Although they all objectively improve on the flaws of the EVM, that’s far from a guarantee of their success. 

Challenges AltVMs face 

With Ethereum dominating the conversation for so long, people have gotten used to a certain way of doing things. The truism that a measure ceases to be a good measure when it becomes a target rings very true for TVL. While on EVM, funds have to be locked to trade; things work very differently on VMs where orderbooks or atomic swaps dominate. 

Once again, the biggest challenge might be the most difficult to deal with: humans. 

“What should change is the perception of users, traders, and marketers to enable a switch to AltVMs.”  - Vitali

Joe shares this sentiment. As a NEAR builder, he is in an ecosystem where the primary dApps are not driven by DeFi or speculation but by what he calls quiet usage.

He also highlights the need for developer education for all altVMs. As a representative of a chain, he asserted that it’s important to ensure that people know what is possible and what isn’t. This includes admitting when a project might be better built elsewhere.

Since the EVM had a headstart, it’s arguably the VM with the most tooling available. Dmitry mentions that, at a basic level, everything works and that retrieving data is a fairly predictable process. With high observability and ossification of EVM and RPC specs, the EVM is still ahead for now. It might take AltVMs 2 - 3 years to reach a similar status. 

The lack of tooling and ease of access to others building on a native VM creates what Solana has termed a developer experience akin to “chewing glass.” 

“For any chain that’s new, the developer experience will be like that, chewing glass.” - Vitali

Any AltVM that fails to attract sufficient developers will ultimately fail. There’s no point in having a chain for the sake of it. Blockchains aren’t like art. 

So far, only one alt VM has really managed to establish itself: Solana. 

“[Solana] seems to be the number three for now, so at least we can say there are more than 2 blockchains”  - Mikko 

Learning from Solana’s success

It’ll be hard for any upcoming alternative VM to engineer the phoenix-like rise of Solana after the FTX fallout. And it might not be all that desirable for crypto's reputation to have another CEX failure and long-dragging lawsuit. Still, there are lessons to be learned from its success in establishing a thriving community. 

Speakers especially pointed out Solana’s hackathon strategy, which took on the existing format of hackathons lasting a max of 3 days and turned it into something different. Solana hackathons run long enough to allow devs to build a functional MVP. The prizes awarded serve as a grant that allows them to take this MVP and turn it into a working business. 

An endeavor that takes up to 18 months, as shared by Vitali from his experience building Spark. 

Regarding how to organize hackathons, Joe recommended that hosting them as part of larger crypto community events has been hugely successful in getting in front of the right people. 

Of course, other factors might have contributed to Solana’s rise, such as its engineering-first approach, the foundation's reluctance to build features only when demanded (like you’d manage a real product), and the way it involved VCs in its ecosystem. Add to that becoming the home for DePIN, at least one bald guy, and memecoins, and you get a solid foundation. 

Will that ensure it stays at the top for the future? Time will tell. 

Hopes for AltVMs

The speakers had different hopes for AltVMs, yet all of them believe that they are a net positive in the Web3 space. After all, what kind of decentralization is it if we’re all forced to use the same VM and the same chain? 

“They provide an opportunity for more devs to have a choice which is a win for everyone. We’ll see what remains.” - Joe

“AltVMs are supposed to let us build better products,” Vitali said, explaining that end-users don’t care about VMs.

Mikko aimed higher: “I hope that when I go to Coingecko eventually, there is something else on spots 1 and 2. Because those deserve to die (Bitcoin & Ethereum)”.  

As so often, a space couldn’t finish without mentioning how early we are and that, hopefully, we can stop talking about infra altogether since it will be taken for granted and as embedded as https. 

Don’t worry, though; at Subsquid, we’ll continue talking about infra until morale consumer crypto adoption improves. And beyond.