Gonka, a decentralized AI computing network, has announced significant updates to its Proof-of-Concept (PoC) mechanism, aiming to align with real computing power and ensure continuous GPU participation. The changes include using a unified large model for both PoC and inference, transitioning to near real-time PoC activation, and optimizing computing power weight calculations to better reflect actual costs. These adjustments are designed to enhance network stability and security as computing power expands, supporting larger AI workloads. Co-founder David emphasized that these updates are not for short-term gains but are essential for evolving the network's consensus and verification structure. The network has reduced PoC activation time to under 5 seconds, minimizing computational waste and improving GPU resource utilization. Additionally, the system now allows single-card and small-to-medium-scale GPUs to earn rewards and participate in governance through flexible mining pool collaboration and inference tasks.