The Render Compute Network is closer than ever to kicking off its US based trial, marking a new phase for decentralized AI infrastructure. Node operators in the US are now onboarding and will soon start supporting AI workloads including inferencing, offline ML, and edge compute.🧵

2/ Since July, we’ve been onboarding US based node operators who have registered and connected their machines to the Render Compute Network. The goal: test real world performance and availability across a range of AI tasks

3/ One early takeaway? The most common GPU model onboarded so far is the NVIDIA GeForce RTX 5090, driven by benchmark results and the geographic distribution of our early customers. Here’s a snapshot (the mix changes depending on when the snapshot is taken) 🟢  = online nodes 🔵 = nodes currently running a job

4/ With more nodes coming online, we are now tracking sessions, giving us visibility into uptime and job activity across the network. Here is a snapshot of node session activity from a 7-day stretch in late July.

5/ Node operators are rewarded in RENDER. They earn availability rewards for simply being online and connected, and additional rewards for completed jobs. Rewards started accruing from their first moment of network availability.

6/ As the trial progresses, we’ll continue to share updates on node growth, GPU distribution, workload types, and more. This is just the beginning.