# MoRI + vLLM: Wide Expert Parallelism and RDMA KV-Cache Transfer for Disaggregated MoE Serving on AMD

> Полный текст доступен на английском

**Published:** 14 октября 2026 г.
**Author:** PyTorch
**Source:** YouTube — PyTorch Official
**Language:** en-US
**HTML:** https://aiqazaqstan.com/ru/articles/mori-vllm-wide-expert-parallelism-and-rdma-kv-cache-transfer-for-disaggr

Hi, I'm Ravi Gupta and AI models researcher. >> Hi folks, my name is Rishi Madhuri. I'm a GPU software developer. >> Hi everyone, I'm Shiksha AI engineer. We all work at AMD. We are going to have a poster session about wide expert parallelism and RDMA KV cache transfer at…

> Материал агрегирован из первоисточника: **YouTube — PyTorch Official**
> [Читать полностью в YouTube — PyTorch Official](https://www.youtube.com/watch?v=cVSkL3dI2b8)

**Теги:** #GPU & Суперкомпьютеры #YouTube

---

*AI Qazaqstan — non-commercial informational portal about AI in Kazakhstan. Machine translations and summaries are AI-generated.*