Skip to content
EthCC - Cannes, France

btchip, Illia Polosukhin, Justin Drake - Battle for Scale: Sharding vs. Rollups

Nicolas Bacca, Illia Polosukhin, Justin Drake

0:000:00

Key takeaways

  • Synchronous Danksharding: Justin Drake says Ethereum has pivoted to synchronous data-availability and execution sharding, branded “danksharding,” letting the base chain host unlimited programmable “native roll-ups” without sacrificing decentralization.
  • Roll-up UX Shortfalls: Illia Polosukhin argues today’s roll-ups fragment liquidity and force wallet hopping; true sharding would make scale an invisible implementation detail for end-users.
  • Security Councils as Liability: Both note that most roll-ups still depend on emergency multisig “security councils,” revealing immature fraud-proof and SNARK tooling; native roll-ups aim to eliminate this crutch.
  • Builder Centralization Inevitable: Drake concedes block construction is a specialized, increasingly centralized business; Ethereum’s answer is Proposer-Builder Separation, MEV-Boost and FOSSIL to block censorship even if only a handful of builders exist.
  • 600× L1 Capacity Target: Drake outlines a plan to triple base-layer throughput every year for six years, turning Ethereum into a “giga-scale Layer 1” that competes with roll-ups on raw performance.
  • L2s as Distributed Sales Force: Roll-up teams function as independent business-development arms, landing integrations with firms like Robinhood and Deutsche Bank—an advantage Ethereum did not foresee in 2018.

On Scaling Architectures

The debate centers on whether roll-ups or sharding delivers a smoother path to mass adoption. Justin Drake positions synchronous danksharding as sharding-in-spirit but roll-up-in-practice. By anchoring data and proofs directly on Layer 1, each “native roll-up” can alter its sequencer design, tokenomics or governance yet keep full Ethereum security. Illia Polosukhin pushes back, calling roll-ups a stopgap that forces users to bridge assets and developers to redeploy contracts across fractured liquidity pools; proper sharding, he says, would hide these complexities entirely.

On Security Model Evolution

Both speakers agree the current roll-up stack is in an awkward adolescence. Emergency security councils exist because fraud-proof games and SNARK circuits can still fail. Drake argues that once proofs are verified in-protocol, councils disappear and roll-ups inherit Layer 1 finality automatically. Polosukhin counters that the ecosystem’s tacit acceptance of multisig control has already shifted the Overton window away from decentralization.

On Block Building & Censorship

Drake acknowledges that high-performance block building centralizes quickly. Ethereum’s roadmap therefore formalizes Proposer-Builder Separation and relies on middleware like MEV-Boost so any validator can outsource block construction without surrendering censorship resistance. Polosukhin notes that sub-second block times leave little room for open competition, making private dark-pool relationships a de facto requirement today.

On Business Development Dynamics

Roll-ups have unintentionally solved Ethereum’s go-to-market problem. Dozens of independent teams court enterprises, each offering tailored economics while still settling on Ethereum. Drake sees this federated BD engine as a strategic asset that future native roll-ups will retain. Polosukhin warns that the appeal rests on centralized performance shortcuts that may undermine the ecosystem’s long-term credibility.

Key actionable insights

  • Track progress on in-protocol proof verification; removal of security councils will be the clearest signal that “native roll-ups” are production-ready.
  • Monitor block-builder concentration and the real-world effectiveness of MEV-mitigation schemes like FOSSIL as throughput rises.
  • Watch whether roll-up user-experience hurdles—bridging, liquidity fragmentation—improve before Ethereum’s projected 600× base-layer expansion is realized.