Uploaded September 2026 | Updated November 2026, 2 days ago
A technical deep dive into vLLM on Intel platforms.
At Ray Summit 2026, Chendi Xue, AI Software Engineer at Intel, introduces Intel's optimized attention, mixture-of-experts, and GEMM operators powered by Triton and SYCL, shows how Intel's next-generation GPUs support native sub-8-bit microscaling formats (MXFP8/MXFP4), demonstrates how decoupling compute-bound prefill and memory-bound decode onto dedicated nodes eliminates inter-token latency spikes, and highlights Intel's backend support for vLLM-Omni for any-to-any multimodal models spanning text, video, and audio.
You'll leave with a map of vLLM performance across Intel GPUs, CPUs, and Gaudi.
This session was part of the first vLLM Conference at Ray Summit, hosted by Inferact.
Liked this video? Check out other Ray Summit vLLM session recordings: youtube.com/playlist?list=PLXMguE8Nc9o4
Subscribe to our YouTube channel to stay up-to-date on the future of AI: youtube.com/c/anyscale
🔗 Connect with us:
LinkedIn: linkedin.com/company/joinanyscale
X: https://x.com/anyscalecompute
Website: anyscale.com
A technical deep dive into vLLM on Intel platforms.
At Ray Summit 2026, Chendi Xue, AI Software Engineer at Intel, introduces Intel's optimized attention, mixture-of-experts, and GEMM operators powered by Triton and SYCL, shows how Intel's next-generation GPUs support native sub-8-bit microscaling formats (MXFP8/MXFP4), demonstrates how decoupling compute-bound prefill and memory-bound decode onto dedicated nodes eliminates inter-token latency spikes, and highlights Intel's backend support for vLLM-Omni for any-to-any multimodal models spanning text, video, and audio.
You'll leave with a map of vLLM performance across Intel GPUs, CPUs, and Gaudi.
This session was part of the first vLLM Conference at Ray Summit, hosted by Inferact.
Liked this video? Check out other Ray Summit vLLM session recordings: youtube.com/playlist?list=PLXMguE8Nc9o4
Subscribe to our YouTube channel to stay up-to-date on the future of AI: youtube.com/c/anyscale
🔗 Connect with us:
LinkedIn: linkedin.com/company/joinanyscale
X: https://x.com/anyscalecompute
Website: anyscale.com










