SOON SVM L2: Can Solana's Virtual Machine Power a Universal Superchain?

Solana is already the fastest general-purpose blockchain in production. Sub-400 millisecond blocks, theoretical throughput of 65,000 transactions per second, fees measured in fractions of a cent. So why would anyone build a Layer 2 on top of it — and why would Solana's own co-founder invest in one?

That question sits at the heart of SOON, a project that is not really trying to scale Solana. Instead, SOON is trying to export Solana's engine to every chain on earth.

The Franchise Model for Blockchain Execution

SOON — short for Solana Optimistic Network — describes itself as a Universal SVM (Solana Virtual Machine) L2. The founding thesis is elegant: the SVM is the most performant smart contract runtime ever built, and there is no reason it should be locked to a single chain.

Think of it like a restaurant franchise. McDonald's does not only operate in one city. The kitchen system — the execution model — is deployed everywhere. SOON's ambition is to do the same with Solana's execution environment: deploy it on Ethereum, on BNB Chain, on Bitcoin L2s, on Cosmos chains, anywhere developers want Solana-speed execution while settling on a different chain's security and liquidity.

The product suite reflects this vision:

• SOON Mainnet: A general-purpose SVM L2 settling on Ethereum with EigenDA as its default data availability layer
• SOON Stack: A modular Rollup-as-a-Service (RaaS) framework enabling one-click deployment of SVM-based L2s on any L1
• InterSOON: A Hyperlane-based cross-chain messaging protocol providing native interoperability between SOON chains, Solana mainnet, TON, and other ecosystems

The first concrete product of the SOON Stack beyond the mainnet is svmBNB — a full SVM L2 running on BNB Chain as its settlement layer with Avail as its DA layer. SOON was the first SVM chain deployed on Avail's stack, and CZ publicly highlighted it — early signal that tier-one ecosystem players are paying attention.

The Decoupled SVM: Why It Is Not Just a Fork of Solana

The natural skeptic's objection is that SOON is just taking Solana's codebase and running it somewhere else. The technical reality is more interesting.

On Solana mainnet, the Transaction Processing Unit (TPU) is tightly coupled to Proof of History (PoH) and Tower BFT consensus. Validators must participate in both consensus and execution simultaneously, which means roughly 80% of Solana's block space is consumed by validator vote transactions — an enormous overhead that every application on Solana subsidizes.

SOON's engineers performed what they call the Decoupled SVM (dSVM): they surgically removed the TPU from Solana's consensus machinery and placed it under rollup node control. The result:

• Vote transactions are eliminated entirely, recovering the DA budget consumed by consensus overhead
• The sequencer controls the TPU rather than a validator committee
• The SVM becomes genuinely portable — it no longer depends on Solana's PoH or its specific consensus model
• Block times drop to 50 milliseconds (versus Solana's ~400ms)
• Throughput scales from 5,000 TPS in current operation up to 600,000 TPS on the roadmap after Firedancer integration

The codebase is open-source at github.com/soonlabs/igloo, giving developers full visibility into the implementation.

This decoupling matters because it is not an incremental optimization. It is a genuine architectural separation that makes the SVM a modular execution layer rather than a feature of one specific chain.

SOON vs. Eclipse: The SVM L2 Derby

SOON is not the only project trying to bring SVM execution to Ethereum settlement. Eclipse launched its mainnet in November 2024 and is the most direct competitor. The comparison reveals meaningfully different technical and strategic bets.

Feature	SOON	Eclipse
Execution	Decoupled SVM	SVM (Sealevel)
Settlement	Ethereum (+ BNB, Bitcoin EVMs)	Ethereum only
Data Availability	Pluggable (EigenDA, Avail, Celestia, ETH blobs)	Celestia only
Fraud Proofs	Optimistic (OP Stack-style)	ZK (RISC-Zero)
Fundraising	$22M via NFT sale	$65M via VC rounds
dApps at launch	~20	60+
Strategic claim	Universal RaaS platform	First mover on SVM+ETH

Eclipse has a meaningful head start in ecosystem depth — more dApps, more capital raised, and a ZK fraud proof system that is technically more trustless than optimistic proofs. SOON's counter is breadth: it is not betting on one chain combination, but on a framework that can be deployed anywhere. The optimistic proof approach also ships faster and is easier to audit, at the cost of the withdrawal challenge period.

Both projects are competing for the same developer audience — Rust developers who know Anchor and Sealevel and want to reach Ethereum liquidity without abandoning the execution environment they know. The race is essentially between Eclipse's depth (more partners, more capital, stronger cryptographic guarantees) and SOON's breadth (more settlement options, more DA options, the SOON Stack as a product platform).

The EVM Parallel and Why It May or May Not Apply

SOON's investors and team frequently invoke the EVM's success as a template. The argument: Ethereum launched the EVM in 2015, and by 2025, more than 50 EVM-compatible chains share the same tooling — Hardhat, Foundry, MetaMask, Etherscan, Solidity. A developer who learns Solidity once can deploy to Ethereum, Polygon, Arbitrum, Base, Avalanche, and dozens of other chains. That tooling network effect is enormously powerful.

The SVM has a parallel opportunity. Approximately 5,000 active developers currently use Rust and Anchor on Solana — a smaller but highly skilled cohort. If SOON Stack succeeds in deploying SVM to Ethereum settlement, BNB Chain, Bitcoin L2s, and Cosmos chains, those developers could suddenly reach the liquidity of every major ecosystem without rewriting their applications.

The case for this parallel is genuine:

• Caldera, the largest RaaS platform (which also deploys OP Stack and Arbitrum Orbit chains), integrated SOON Stack — a meaningful distribution channel
• Solana Foundation chair Lily Liu and co-founder Anatoly Yakovenko both invested in the $22M NFT sale, providing ecosystem credibility
• Mustafa Al-Bassam (Celestia co-founder) and Robinson Burkey (Avail co-founder, Wormhole co-founder) are also angels — DA layer relationships are built into the cap table
• svmBNB is live and running, meaning the multi-settlement thesis has at least one production proof point beyond SOON Mainnet itself

The case against is also real:

• The EVM's success was driven by Ethereum's $40B+ DeFi TVL acting as gravitational pull for new chains. SVM chains can tap into Ethereum's liquidity via bridges, but they do not natively inherit it
• Rust + Anchor is meaningfully harder than Solidity for average developers, slowing the developer network effect
• SOON's current TVL is in the $6.5M–$15M range — small relative to Eclipse and tiny relative to the EVM ecosystem
• The SOON token's path from an ATH of \4.87 in November 2025 to roughly $0.16 by mid-2026 reflects that the market has not yet validated the long-term thesis

Firedancer: The Wildcard

One of the more counterintuitive elements of SOON's roadmap is its explicit dependency on Firedancer — Jump Trading's independent Solana validator client that is expected to push Solana mainnet to ~650,000 TPS.

SOON plans to integrate Firedancer into the SOON Stack, meaning SOON chains would inherit Firedancer's throughput improvements as they land. The irony: a project building L2s on top of Solana is betting that Solana itself gets dramatically faster, and plans to use that improvement in its own execution layer.

This is actually a coherent strategy. If Firedancer makes Solana too fast for anyone to credibly argue it needs L2s, SOON's answer is: SOON is not scaling Solana, it is exporting Solana's engine. The performance ceiling still rises because SOON chains run the same SVM code.

Firedancer reached partial deployment on Solana mainnet in 2024-2025 with full integration ongoing. As it rolls out, SOON Stack chains should receive a significant throughput upgrade without requiring any application-level changes.

Who Actually Builds on SOON?

The honest answer in mid-2026 is: early-stage adopters and ecosystem experimenters.

The partnership roster includes UXLINK (one of the largest Web3 social platforms globally, with 20 million+ users), Wormhole for cross-chain asset transfers, and the Avail DA team. The 20+ projects on SOON Mainnet at public launch represent a reasonable start for a new L2 framework, but nowhere near the density needed to claim ecosystem success.

SOON's "community-driven" positioning — raising $22M via NFT sale rather than traditional VC rounds, with the stated principle of "zero VC privileges" — resonates with certain Web3 communities and distinguishes it from Eclipse's more conventional institutional fundraise. Whether that positioning translates into actual developer and user loyalty is an open question.

The SOON Stack's RaaS offering is arguably the most important long-term product. If game studios, high-frequency DeFi protocols, and AI agent coordination networks want dedicated, sovereignty-preserving chains with SVM execution, the one-click deployment model positions SOON as the infrastructure provider rather than competing for liquidity on a single chain.

The Honest Assessment

SOON represents a coherent technical thesis executed by a credentialed team with genuine ecosystem relationships. The Decoupled SVM is a real architectural innovation, not just rebranding. The multi-settlement, pluggable-DA approach gives it more adaptability than single-stack competitors.

The headwinds are also real. Eclipse has a head start. Solana's own performance improvements reduce the urgency of L2 scaling. The token has experienced severe drawdown. And the EVM parallel, while inspiring, took years and tens of billions in Ethereum DeFi liquidity to materialize.

The project to watch is whether the SOON Stack — not SOON Mainnet specifically — gains traction as a RaaS platform. If developers start deploying SVM chains on BNB, Bitcoin L2s, and Cosmos chains via SOON Stack tooling, the universal SVM thesis becomes self-reinforcing. If SOON Mainnet remains the only live SVM+Ethereum chain under this brand while Eclipse captures the Ethereum-specific market, SOON's moat shrinks to the multi-settlement niche.

The SVM portability race is genuinely interesting because the technical merits are real. The ~1.4 million active addresses on SOON and a handful of live chains are the first data points — not the verdict. Whether those merits translate into the kind of developer network effect that made the EVM ubiquitous — that remains the billion-dollar question, and the answer will likely be visible in the SOON Stack deployment count twelve months from now.

BlockEden.xyz provides enterprise-grade Solana RPC and API infrastructure for developers building on SVM-based networks. Explore our API marketplace to access high-performance endpoints designed for production workloads.