Skip to main content

Zero-Knowledge Rollups

There are a variety of robust SDKs available today that facilitate the deployment and management of sovereign blockchains. However, when working with ZK-rollup ecosystems, one critical component is the maintenance of your own provers. This can often be one of the most challenging and non-negotiable aspects of the stack.

Key challenges you may encounter include:

GPU provisioning​

If you reserve a fixed number of GPU instances, your chain can only progress at a fixed rate. You are over-paying in times of lower TPS and delaying finality in times of higher TPS. On the other hand, scaling your number of GPU provers to meet the needs of your chain requires mature DevOps expertise and monitoring.

Version control​

Staying up to date with cryptographic protocol changes is often mandatory to participate within the broader ecosystem. These changes may take place either:

  • within the proving software
    • changing the nature of your hardware requirements,
  • within circuit definitions at the zkEVM level
    • requiring you to update your circuit artifact files.

Observability and troubleshooting​

Monitoring the progress of zkEVM proofs and troubleshooting in case of hardware failures or edge cases requires hard-won subject matter expertise. Sindri's front end provides users with real time progress updates on proof generation. Our on-call staff can provide guidance in case of misconfiguration between a chain and provers.

If you are interested in deploying your own Layer 2 or Layer 3 solution but want to avoid the complexities of managing proof generation infrastructure, you can find more information on Sindri's integrations below.