High-performance blockchains and shared state

Nov 9, 2025

There is a lot of excitement around the new cohort of L1/L2s promising groundbreaking performance. We are entering the real-time era of blockchains and faster, cheaper compute is one of those things that “is not going to change in the next 10 years”. Users will always value faster and cheaper transactions. However, performance increase alone is not enough for these blockchains to be successful. Even if high-performance L1s/L2s deliver on their promises, they still face the massive challenge of bootstrapping an ecosystem from scratch—attracting developers, users, liquidity, and security. It’s relatively easy to build a hyper-fast blockchain with centralized assumptions. The tricky part is how to bootstrap the biggest ecosystem if you are an L1 and/or be composable with the base layer if you are an L2, while not giving up on trust assumptions. Even assuming similar security properties, the performance gains of these networks come with the huge cost of fragmentation. I’d argue they can strive to become siloed ecosystems at best, and ghostchains at worst.

Shared state is why Solana is winning Composability is one of the main ingredients of Solana’s success. There are many L3s that are faster than Solana, but they are relatively shitty products because there is nothing to do there. They are completely detached from the rest of the world where stuff happens. A single-state machine is multiplicative. If someone deploys a new program on Solana, existing programs can interact with it. A simple example of this is the token program. All new tokens on Solana are instances of the same token program implementation, reducing the risks of rugs, allowing you to trade the new token on Jupiter and visualize it on Phantom or Backpack as expected. No bridging, no need for multiple versions of the same token. There is only one version of native USDC that everyone uses. Things are simpler and compound faster because every new token adds value to the whole ecosystem. Shared state ultimately means stronger network effects and less friction for users and developers.

Composability means better products. Seeking performance optimizations without thinking of composability and from an ecosystem perspective inevitably leads to fragmentation and poor UX because blockchains are isolated silos. Now all of a sudden users have to bridge. The slippage for USDC/SOL is a few bips higher. That new memecoin is not showing up in the wallet anymore. Whether it’s a new, shiny L1 or L2, the product just got worse.

The fall of the Giant This should be a relatively uncontroversial take, given it is exactly what happened to Ethereum with the rollup-centric roadmap. In a market-induced race for cheaper block space and faster block time, >50 “aligned” teams spun up an equal number of ecosystems with different tradeoffs, standards, and communities, all competing with each other for developers and attention. It is the exact opposite of network effects and a multiplicative ecosystem. This dispersion of incentives is the fundamental reason why Ethereum is struggling right now and losing market share to Solana. The risk we face is witnessing history repeat itself with the SVM now that Solana is leading the charge. While, on one hand, more forks imply that the SVM becomes the new industry standard—which is beneficial for attracting more developers—it also poses a potential threat if we forget what brought us here in the first place. Market forces will push many teams to explore launching their own version of Hyperliquid, to build the ‘OP stack of SVM,’ and to launch their appchain, etc. Because these networks are fundamentally disconnected and compete through BD efforts, token grants for developers, and user incentives, the logical end state is a bunch of half-cooked ecosystems producing products strictly worse than if they had been built on Solana in the first place.

Solana as the single, unified ecosystem The success of Solana as a single, unified ecosystem is thus imperative for the success of our industry. For Web3 to succeed and legitimize itself, we must build better products than Web2 solutions. We need more performance, more scale AND a liquidity layer where the state lives and programs compose seamlessly. We need to collectively reject the idea that speed and composability are mutually exclusive. Instead, we must double down on technologies and platforms that deliver both. Having just one or the other is simply not enough to win. Web3 is about breaking barriers and giving more access, not rebuilding silos. Solana embodies that vision, and at MagicBlock, we’re here to help make it a reality.