Matter Labs’ Airbender Prime Prover Processes 99% of Ethereum Blocks in Under 10 Seconds

Matter Labs has announced the arrival of Airbender Prime, a zero-knowledge prover that the company claims successfully processed 99% of 7,200 consecutive Ethereum mainnet blocks in under 10 seconds each. This specific benchmark was executed using a single NVIDIA RTX PRO 6000 Blackwell GPU, a hardware configuration that intentionally distinguishes this particular test from more common multi-node or cloud-based verification setups.

Benchmark Specs and Hardware

The core of this claim rests upon two very specific metrics: the total percentage of blocks that were successfully proven and the exact time taken per individual block. Matter Labs stated that Airbender Prime handled 99% of the 7,200 blocks that were tested. In practical terms, this means that for approximately 7,128 blocks, the prover was able to complete its required work within the 10-second limit. The remaining 1% of blocks within the sample set did exceed that threshold, although the company did not specify the exact duration for those specific outliers in the initial announcement.

The hardware utilized for this test was a single NVIDIA RTX PRO 6000 Blackwell GPU. This detail is significant because many high-performance computing tasks within the realm of cryptography typically require distributed systems or the use of specialized accelerator clusters. By demonstrating that a single, commercially available workstation-class GPU can handle the vast majority of Ethereum mainnet block proofs quickly, Matter Labs suggests there is a lower barrier to entry for nodes wishing to participate in the verification process. The RTX PRO 6000 Blackwell is part of NVIDIA’s newer architecture, which is designed specifically for professional workstations, indicating that the prover is optimized for modern consumer and prosumer hardware rather than being restricted to enterprise data centers.

Ethereum mainnet blocks are typically produced every 12 seconds. A prover that can generate a validity proof in under 10 seconds operates comfortably within the block time window, which is critical for maintaining real-time finality. If the proof generation time were to consistently exceed the block interval, the layer-2 network would inevitably face backlogs or significant delays in confirming transactions. The 99% success rate within 10 seconds indicates that the system is capable of keeping pace with the base layer under normal operating conditions.

Starknet’s L1 Consideration

The ZkSync announcement occurred five days after Starknet made its own strategic statement. Starknet, another leading zk-rollup project, disclosed that it is actively considering the possibility of becoming an independent layer-1 (L1) blockchain. This represents a distinct architectural path from the layer-2 model that both ZkSync and Starknet currently operate under. A layer-2 network relies on Ethereum for its security and settlement, while a layer-1 blockchain is self-sovereign, generating its own security and its own finality.

Starknet stated that this potential transition is targeting full quantum resistance. Quantum computing poses a theoretical threat to many current cryptographic algorithms, including elliptic curve cryptography, which is used extensively in blockchain signatures and key management. By moving to an L1 with a quantum-resistant design, Starknet aims to future-proof its network against potential advances in quantum technology. This contrasts with the immediate focus of ZkSync on prover efficiency and hardware optimization, though both strategies address different aspects of long-term network viability.

The timing of these announcements places two major players in the zk-rollup space in a competitive narrative, albeit with different focal points. Matter Labs is emphasizing performance metrics and hardware accessibility, while Starknet is highlighting architectural independence and cryptographic longevity. Both developments reflect the maturing phase of layer-2 technology, where competition is shifting from basic functionality to specialized capabilities like speed, cost-efficiency, and security against emerging threats.

The 7,200-block test represents a sample size of roughly 24 hours of Ethereum mainnet activity, given the 12-second block time. This duration is sufficient to capture typical network conditions, though it may not account for extreme congestion or specific edge cases that could affect prover performance. The 99% success rate suggests high reliability, but the absence of detailed breakdowns for the remaining 1% leaves some questions about the consistency of the performance across all possible block states.

Implications for Rollup Infrastructure

For ZkSync users and developers, the Airbender Prime benchmark suggests that the network can maintain low-latency finality without requiring expensive, multi-GPU setups for standard operations. This could potentially reduce the operational costs for nodes and increase the decentralization of the prover network, as more participants may be able to afford the necessary hardware. The use of the NVIDIA RTX PRO 6000 Blackwell specifically indicates that the prover is compatible with the latest generation of GPU technology, which may offer improved performance-to-watt ratios compared to older models.

The mention of Ethereum mainnet blocks is crucial because it ties the layer-2 performance directly to the base layer’s cadence. Any improvement in prover speed that allows it to stay under the block time is a direct enhancement to the user experience, as it minimizes the waiting time for transaction confirmations. The 10-second threshold is not arbitrary; it aligns with the natural rhythm of Ethereum’s proof-of-stake consensus mechanism. By meeting this threshold for 99% of blocks, Matter Labs demonstrates that the Airbender Prime prover is robust enough for production environments.

Starknet’s consideration of an L1 transition adds a layer of complexity to the zk-rollup ecosystem. If Starknet proceeds with this change, it would no longer rely on Ethereum for data availability and security, which could impact its relationship with the broader Ethereum ecosystem. This move is often driven by the desire for greater control over transaction fees, block times, and governance. However, it also means that Starknet users would no longer have the security backstop of Ethereum’s massive hash rate, making the quantum-resistant design a critical component of its new value proposition.

Matter Labs did not provide additional details on the specific cryptographic algorithms used in Airbender Prime or whether it incorporates quantum-resistant techniques. The focus remains on the speed and hardware efficiency of the prover. This distinction is important for investors and developers evaluating the long-term security posture of different rollup solutions. While speed is a critical feature for user adoption, cryptographic resilience is a foundational requirement for any blockchain that aims to secure assets over a multi-decade horizon.

The five-day gap between the two announcements suggests a coordinated or reactive dynamic in the market, though no direct causal link was stated by either company. Both projects are addressing different pain points in the current layer-2 landscape: Matter Labs is tackling the bottleneck of proof generation speed, while Starknet is addressing the potential existential threat of quantum computing and the limitations of layer-2 dependency.

For the Ethereum ecosystem, these developments signal continued innovation in the rollup space. As layer-2 networks compete for users, they are increasingly differentiating themselves through technical metrics and strategic architectural choices. The ability to process proofs quickly on accessible hardware is a tangible benefit for developers and users, while the exploration of quantum-resistant L1s represents a long-term bet on the future of cryptographic security.

There are no immediate changes to the ZkSync or Starknet networks resulting from these announcements. The Airbender Prime prover is a new tool within the ZkSync infrastructure, and its deployment will likely be gradual. Starknet’s L1 consideration is a strategic option, not a confirmed decision, and any transition would require significant development work and community consensus. The current state of both networks remains unchanged, with ongoing operations on the Ethereum layer-2 standard.

Further details on the Airbender Prime prover, including source code availability and third-party audits, have not been disclosed in the initial statement. The 99% success rate is based on Matter Labs’ internal testing, and independent verification of these benchmarks is pending. For Starknet, the roadmap for a potential L1 transition, including timelines and technical specifications, has not been released. The focus for now remains on the stated intent to target full quantum resistance.