🔵 𝗠𝗮𝘀𝘀𝗶𝘃𝗲 𝗻𝗲𝘄𝘀 in the AI infra space!  The @sgl_project creators officially launched their startup, @radixark, with a $𝟭𝟬𝟬𝗠 Seed round at a $𝟰𝟬𝟬𝗠 𝘃𝗮𝗹𝘂𝗮𝘁𝗶𝗼𝗻. I was lucky to catch up with their Co-founder & team (@liin1211) back at #RaySummit to ask what they were cooking to redefine how frontier AI is built and scaled 🔥! Big congrats to the SGLang team and @ying11231 Announcement👉🏻: lnkd.in/ebsjP8Ma #SGLang #RadixArk #SeedRound #AIInference

GPU memory alone won’t carry the next generation of LLM serving. At #RaySummit, our Chief Scientist @this_will_echo shared how #LMCache offloads KV Cache across CPU RAM, local disk, Redis, and S3, while enabling cache reuse beyond basic prefix caching. Watch the full talk on YouTube: 👉🏻youtube.com/watch?v=aVpkkVp-… #RaySummit #LMCache #Tensormesh #KVCache

🚀 #RaySummit 2025 Day 1 Morning — Distributed LLM & RL Systems Ecosystem for scalable, interactive, and resilient LLM/RL systems — merging research agility with production-grade reliability. Scalable, fault-tolerant AI sys across training, RL, agentic. 👉🏻 linkedin.com/posts/junfan-zh…

🚀 #RaySummit 2025 Day 2 Morning Highlights — Physical Intelligence, Code RL, Scalable Inference Systems 💡Robotics, code agents, forecasting, and hardware optimization — unifying data, control, and computation under one distributed paradigm. 👉🏻 linkedin.com/posts/junfan-zh…

Talked to a lot of fantastic people at the #RaySummit ! What a great event! It is always great to learn that SkyPilot solves issues that many teams are dealing with — running & managing AI on multiple clusters/clouds. Thanks for all your interest! Let’s work together to make AI infra better for all!

What an amazing #RaySummit! 100s of conversations revealed the same theme: Managing multiple Slurm/KubeRay/Kueue deployments → more ops complexity, lower GPU utilization. SkyPilot simplifies it → one system to manage every cluster and GPU cloud. 🔗 Give it a try: github.com/skypilot-org/skyp…

Ray Summit 2025 has been exceptional. This is where 1,000 building production AI infrastructure come together. Not just talking about the future. Building it. Open source contributors, ML engineers running systems at scale, and infrastructure leaders sharing real solutions to real problems. @simon_mo_'s talk on @vllm_project was my favorite, including deep dives into high-performance inference, prefix caching, large-scale orchestration, and more. Don't sleep on this impactful project. @thehousefund is a proud first-round backer of @anyscalecompute, who powers this summit. Special to be part of this community pushing the boundaries of distributed computing and AI systems. This is a can't-miss Summit every year. Well done @KeertiMelkote, @robertnishihara, @istoica05, Philipp Moritz & the Ray ecosystem at large! #RaySummit #Anyscale #AI #Infrastructure #OpenSource #vLLM

We’re excited to sponsor @anyscalecompute's #RaySummit ⭐ Today, Dr. Jim Fan took part in the keynote speech, and we connected with so many talented AI/ML builders to share hands-on insights. Explore how the NVIDIA Developer Program is driving open-source innovation and powering the future of AI ➡️ nvda.ws/4ophYLv

Great talk on @datologyai trillion token synthetic data pipeline by @hurrycane and Fan Pan at #raysummit!

Catch Exosphere today at Ray Summit 2025. If you are interested in very large scale AI agents feel free to ping, happy to chat at the summit. #raysummit

the amount of aura on stage right now is insane #raysummit - founding eng of kubernetes (dawn chen, no x found) - cofounder of anyscale @robertnishihara - lead of vllm @simon_mo_ - ex-lead at pytorch @joespeez

🚦Green light: win an NVIDIA DGX Spark (~$4,500 AI supercomputer). 🛑Red light: missing us at #RaySummit. Find the hot-pink Squid Game crew, say hi, and enter the raffle. See you tomorrow! 🦑💖

We are happy to release SkyRL tx 0.1 novasky-ai.notion.site/skyrl…, an open source unified training and inference engine that supports the Tinker API. This release has many performance enhancements and also new features but most importantly RL training is now working end-to-end. If you are interested in the project and are coming to #RaySummit, we are giving a talk about SkyRL tx tomorrow (Nov 4) at 4pm, come join us!

If you're heading to #RaySummit, stop by booth T4 to chat with our team about running, managing, and scaling AI workloads across any infrastructure.

Your infra can’t fix what it can’t see 👀 At #RaySummit 2025, Mengjin Yan & Nikita Vemuri from @anyscalecompute reveal Ray-native observability – dashboards built for distributed AI, storing telemetry in your own cloud Plus: a new Ray Export API to extend monitoring to your stack

Our CTO, Devansh Ghatak will take the stage at #RaySummit 2025 in San Francisco! 🚀 He's diving deep into "Orchestrating the GenAI Lifecycle with KubeRay," showcasing how to simplify distributed training, fine-tuning, and inference on Kubernetes. 📅 12:30 PM | Nov 4 #GenAI

Our talk at #RaySummit is next week on Tuesday 11/04 at 4pm! We’ll be speaking about how @datologyai scales synthetic data generation to trillion token scale and the engineering lessons around it! See you there!

See what changed and why 🔎 At #RaySummit 2025, Elizabeth Hu, Goku Mohandes & Akshay Malik from @anyscalecompute will announce Lineage Tracking on Anyscale: end-to-end ML lineage powered by OpenLineage. Nov 3–5 | SF Register: na2.hubs.ly/H01Q0LS0

Infra that ships. Ideas that scale. Keynotes at #RaySummit 2025: • Chelsea Finn (robot learning, meta-learning) • Jimmy Ba (Adam optimizer, DL/RL) • Peter Ludwig (safe AI at scale) Nov 3–5 SF Save your seat: 🔗na2.hubs.ly/H01MYqw0

Thanks for all people dropping by our @skypilot_org booth! Really enjoyed the convos and good to learn that we are resolving urgent pains companies have: run and manage AI on any GPU cluster, no matter if it is reserved, on-prem, or from clouds. See you all at #RaySummit ; )