We’re excited to announce a new partnership with Lilypad. This collaboration will combine Lilypad’s on-demand GPU compute with the Nexus network, enabling secure and verifiable computation around the world.

GPU compute for a verifiable network

Lilypad provides on-demand GPU compute through a decentralized marketplace, making high-performance computation accessible for tasks like ZK proofs, AI inference, and large language model fine-tuning.

By integrating Lilypad’s GPU resources with Nexus’s network, developers can achieve scalable, cost-effective, and verifiable computation for distributed networks.

What this means for developers

Integrating the Nexus network with Lilypad’s GPU network unlocks new possibilities:

• Verifiable GPU compute: Developers can execute GPU-intensive  tasks like AI inference or bioscience modeling with confidence, knowing computations are cryptographically verified.
• Privacy-preserving applications: Build apps that protect sensitive data while ensuring the correctness of computation.
• Decentralized infrastructure: Access compute resources without relying on centralized providers, reducing costs and opening up opportunities to run any kind of model, securely, on a network of available infrastructure.

Unlocking new use cases

With this partnership, developers can create applications that combine secure computation and verifiable results, such as:

• Identity verification: Verify identity without exposing personal data.
• AI + ZK research: Train and deploy AI models with cryptographic guarantees.
• Privacy-preserving analytics: Perform computation on private datasets without revealing underlying information.

The future of secure compute

We recently completed a testnet, enabling users to explore verifiable computation at scale.

Together, Lilypad and Nexus are making secure, decentralized compute a reality — laying the groundwork for a more secure and Verifiable Internet.

Stay tuned as we roll out testnets and explore how the Nexus network and Lilypad’s GPU network can work together to power the future of computation.