# The future of zero-knowledge rollups By [SwapSpace](https://paragraph.com/@swapspace) · 2025-02-06 --- Zero-Knowledge Rollups are at the forefront of Ethereum’s scaling revolution, offering trustless verification, lower fees, and enhanced privacy. As zkEVMs, recursive proofs, and decentralized proving networks evolve, ZK-Rollups are ready to redefine blockchain efficiency. But will they fully replace Optimistic Rollups? This article by SwapSpace CEO Andrew Wind explores the future of ZK-Rollups, their challenges, and the innovations shaping their adoption. The state of ZK-Rollups today ============================= Zero-knowledge rollups or ZK-Rollups have rapidly become one of Ethereum's most promising scaling solutions, significantly reducing transaction costs and improving throughput without compromising security. By using zero-knowledge proofs, these rollups bundle multiple transactions into a single proof, which is then verified on-chain. This allows Ethereum to handle up to 100,000 transactions per second (TPS) while maintaining its security guarantees. Currently, several major projects are leading the ZK-Rollup space, including [StarkNet](https://www.starknet.io/), [zkSync](https://zksync.io/), [Polygon zkEVM](https://polygon.technology/polygon-zkevm), and [Scroll](https://scroll.io/). Each has a unique approach to proof generation, decentralization, and Ethereum Virtual Machine (EVM) compatibility. **_Important!_** _While StarkNet uses STARKs (Scalable Transparent Argument of Knowledge) for its proofs, zkSync and Polygon zkEVM focus on making ZK-Rollups fully EVM-compatible, enabling seamless migration of existing smart contracts_. However, ZK-Rollups still have some challenges. Generating zero-knowledge proofs is computationally expensive, leading to latency issues and centralization concerns, as only a few entities currently have the capability to generate proofs efficiently. Additionally, developer adoption remains a problem, as writing smart contracts optimized for ZK execution requires specialized knowledge. **_Interesting fact!_** _In 2023, the first-ever recursive ZK-proof was successfully generated on Ethereum, proving that multiple transactions can be aggregated into a single proof, exponentially increasing scalability potential._ The next evolution: recursive proofs & zkEVM ============================================ As ZK-Rollups evolve, two major innovations, such as recursive proofs and zkEVMs are shaping the future of Ethereum scalability. Recursive proofs: scaling without limits ---------------------------------------- Recursive proofs allow multiple zero-knowledge proofs to be combined into a single proof, drastically reducing verification costs on Ethereum. For instance, verifying a single ZK-SNARK proof on Ethereum consumes approximately [500,000 gas](https://ethereum.org/sw/zero-knowledge-proofs/), which, at a gas price of 30 Gwei and an ETH price of $3,000, equates to about $20 per verification. This means rollups can process and aggregate transactions more efficiently before submitting them on-chain. Thus, recursive proofs reduce on-chain gas costs by compressing multiple proofs into one, enable exponential scalability by stacking multiple proof layers and lower latency, making ZK-Rollups more competitive with Optimistic Rollups. Notable Implementations of recursive proofs include: * **Plonky2 (by Succinct Labs):** A highly efficient recursive proof system optimized for speed. * **StarkWare’s SHARP:** Aggregates transactions across different dApps, reducing costs for users. zkEVMs: making Ethereum fully ZK-compatible ------------------------------------------- A major challenge for ZK-Rollups has been EVM compatibility. zkEVMs solve this by enabling smart contracts to run in a zero-knowledge environment without modification. Leading zkEVM projects are: * **Polygon zkEVM & Scroll:** Prioritize efficiency while maintaining high compatibility. * **zkSync Era:** Focuses on developer-friendly tooling and native account abstraction. * **Taiko (Type 1 zkEVM):** Strives for full Ethereum equivalence but faces higher proving costs. Each zkEVM design balances trade-offs between efficiency, security, and developer experience. As zkEVMs mature, they are set to accelerate Ethereum’s transition to a highly scalable, trustless ecosystem. Decentralizing the prover network ================================= One of the biggest challenges in ZK-Rollups today is the centralization of proof generation. Currently, most ZK-Rollup projects rely on a small set of trusted entities to generate and submit proofs to Ethereum. This creates potential problems, censorship risks, and trust assumptions – contradicting the decentralized idea behind blockchain technology. Why decentralization matters ---------------------------- * Reduces reliance on a single entity, making rollups more censorship-resistant. * Improves network security by distributing proof generation. * Lowers costs over time by increasing competition among provers. Approaches to decentralizing provers ------------------------------------ * **Multi-prover networks:** StarkNet is working toward a [decentralized network of provers](https://www.starknet.io/blog/starknet-decentralization-a-roadmap-in-broad-strokes/), allowing multiple participants to generate and validate proofs. * **SNARK-as-a-Service:** Initiatives like Aleo and Mina offer proof marketplaces where different entities can generate ZK-proofs, reducing dependency on centralized nodes. * **Specialized ZK hardware:** Companies like Nvidia are developing [ZK-accelerated chips](https://www.ingonyama.com/blog/revisiting-paradigm-hardware-acceleration-for-zero-knowledge-proofs) to lower computational costs, enabling more participants to act as provers. ZK-Rollups vs. Optimistic Rollups: long-term outlook ==================================================== Ethereum’s Layer 2 ecosystem is currently divided between Optimistic Rollups (ORs) and Zero-Knowledge Rollups (ZK-Rollups), each offering unique advantages and trade-offs. While both solutions scale Ethereum by executing transactions off-chain and posting compressed data on-chain, their security models and finality mechanisms set them apart. * **Optimistic Rollups**, such as [Arbitrum](https://swapspace.co/exchange/arb?utm_source=Mirror&utm_medium=content-marketing&utm_campaign=the-future-of-zero-knowledge-rollups&utm_content=Arbitrum) and [Optimism](https://swapspace.co/exchange/op?utm_source=Mirror&utm_medium=content-marketing&utm_campaign=the-future-of-zero-knowledge-rollups&utm_content=Optimism), assume transactions are valid by default and rely on a fraud-proof system. If an invalid transaction is detected, a challenge period (typically seven days) allows for dispute resolution. This approach makes ORs easy to implement and cost-effective in the short term, but the delayed finality and reliance on honest actors introduce security trade-offs. * **ZK-Rollups**, like [StarkNet](https://swapspace.co/exchange/stark?utm_source=Mirror&utm_medium=content-marketing&utm_campaign=the-future-of-zero-knowledge-rollups&utm_content=StarkNet), [zkSync](https://swapspace.co/exchange/zksync?utm_source=Mirror&utm_medium=content-marketing&utm_campaign=the-future-of-zero-knowledge-rollups&utm_content=zkSync), and Polygon zkEVM, take a different approach by using validity proofs to cryptographically verify transactions before posting them on-chain. This eliminates the need for a challenge period, offering instant finality and enhanced security. However, generating zero-knowledge proofs remains computationally expensive, slowing adoption. Despite Optimistic Rollups currently leading in adoption, Ethereum’s co-founder Vitalik Buterin has predicted that ZK-Rollups will ultimately become [the dominant scaling solution](https://www.theblock.co/post/162098/zk-rollups-likely-to-be-main-layer-2-solution-for-ethereum-says-vitalik-buterin) due to their stronger security guarantees and faster transaction finality. Future innovations and open challenges ====================================== ZK-Rollups are rapidly evolving, but several innovations and challenges will determine their long-term impact. One key area of development is privacy-preserving applications. While ZK-Rollups already enhance scalability, integrating zero-knowledge proofs for on-chain privacy could enable private DeFi transactions, confidential voting, and secure identity solutions. **_Example!_** _Projects like_ [_Aztec Network_](https://aztec.network/) _are actively working on private smart contracts using ZK technology._ Another major innovation is reducing the computational costs of proof generation. Today, generating ZK-proofs is expensive and requires specialized hardware and substantial computing power. Companies like Ingonyama are developing ZK-accelerated hardware, while frameworks like Nova aim to make proof generation significantly more efficient. These advances will be critical for making ZK-Rollups widely accessible. Interoperability is also a pressing issue. Currently, most ZK-Rollups operate as isolated ecosystems, limiting liquidity and composability. Efforts like [zkBridge](https://www.zkbridge.com/) by Polyhedra Network and [Succinct Labs' ZK](https://www.succinct.xyz/) light clients are focused on enabling seamless communication between different blockchains using ZK proofs. However, regulatory uncertainty is still a challenge. Governments are starting to closely examine privacy-focused technologies, which could affect how widely ZK-Rollups are adopted. The future of ZK-Rollups will depend on solving these challenges while continuing to improve scalability and decentralization. Conclusion ========== ZK-Rollups hold the potential to transform Ethereum with enhanced scalability, security, and privacy features. Despite challenges like high proving costs and decentralization, advances in recursive proofs, zkEVMs, and ZK-accelerated hardware are pushing the technology forward. While Optimistic Rollups currently dominate, Buterin predicts ZK-Rollups will eventually take the lead. Overcoming technical and regulatory hurdles will be key to unlocking their full potential, offering a more scalable and decentralized blockchain future. --- *Originally published on [SwapSpace](https://paragraph.com/@swapspace/the-future-of-zero-knowledge-rollups)*