🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

We are excited to share a preview of our peer-to-peer decentralized inference stack Engineered for consumer GPUs and high-latency networks — plus a research roadmap to scale it to a planetary-scale decentralized inference engine.

Demo of 4x4090s running QwQ-30B across 4 different locations

We will integrate this work into our protocol to power our next synthetic data run — onboarding the largest amount of compute yet. Next milestone: SYNTHETIC-2 — a large scale reasoning corpus powered entirely by community GPUs. primeintellect.ai/blog/inference

we have a few open-source infra releases dropping this week. starting today with Mika Senghaas's great work on decentralized pipeline parallel inference

wrote a little something on our learnings from decentralizing inference and open-sourced 3 research codebases. tl;dr optimizing inference under decentralized constraints is worthwhile, non-trivial, and far from solved. excited to be building this with the team! more soon, when we

We're taking a big step towards truly decentralized inference — unlocking consumer GPUs and already outperforming traditional approaches that stall in high-latency settings. Unlike other p2p inference engines (e.g., Petals, Exo), our stack uniquely leverages vLLM’s advanced

Prime intellect ensuring we’re in the good timeline

Very excited to soon release SYNTHETIC-2, partially powered by consumer grade GPUs. Very confident that what we’ve planned for this dataset will be incredibly useful for the open source community

L. F. G.

🕺💃

We wrote an extensive blog post on large-scale pipelined inference and released a VLLM integration to connect any machines over the internet to serve a model. Will be the foundation of our next synthetic-2 run (and later allow consumer GPU to join rl run)

Thanks to compute from Prime Intellect, each of our teams will have access to a dedicated H200. Train whatever you want, GPUs will go brrr. Happening next weekend in Zurich🇨🇭, at the mimic x Loki Robotics x OpenAI Hack

We did it — the first decentralized RL training of a 32B model is complete! Full open-source release is coming in ~1 week, including: checkpoints, data and a detailed technical report.

scaling decentralized reinforcement learning 🫡

Decentralized AI is coming. LFG 🔥🚀

We have 8x h100s spot instances for $1/hr each 🫡 app.primeintellect.ai/dashboard/crea…