# \#vllm
**URL:** https://devcommunity.amd.com/tag/vllm/84.md
[Latest](https://devcommunity.amd.com/latest.md) · [Categories](https://devcommunity.amd.com/categories.md) · [Tags](https://devcommunity.amd.com/tags.md)
---
## [Hyperloom: A Multi-Agent Harness for Autonomous Inference Optimization on AMD GPUs](https://devcommunity.amd.com/t/hyperloom-a-multi-agent-harness-for-autonomous-inference-optimization-on-amd-gpus/1109)
**Author:** [@nirth\_AMD](https://devcommunity.amd.com/u/nirth_AMD)\
**Replies:** 2\
**Last updated:** [October 1, 2026, 9:23am UTC](https://devcommunity.amd.com/t/hyperloom-a-multi-agent-harness-for-autonomous-inference-optimization-on-amd-gpus/1109 "2026-10-01T09:23:29Z")
What’s in it: Hyperloom runs the full optimization loop end to end - serving configuration, kernel-level tuning, a validation sweep across operating points, and preservation of whatever wins - with no human steering it. …
---
## [veRL on AMD: Production-Ready RL Post-Training on ROCm](https://devcommunity.amd.com/t/verl-on-amd-production-ready-rl-post-training-on-rocm/1025)
**Author:** [@nirth\_AMD](https://devcommunity.amd.com/u/nirth_AMD)\
**Replies:** 0\
**Last updated:** [September 14, 2026, 2:35pm UTC](https://devcommunity.amd.com/t/verl-on-amd-production-ready-rl-post-training-on-rocm/1025 "2026-09-14T14:35:20Z")
Reinforcement learning post-training on AMD Instinct GPUs is here — with a turnkey container, AITER-accelerated vLLM and SGLang rollout, and accuracy validated on both MI300 and MI355. veRL — the open-source implement…
---
## [Optimizing ATOM and vLLM-ATOM for High-Interactivity Inference](https://devcommunity.amd.com/t/optimizing-atom-and-vllm-atom-for-high-interactivity-inference/944)
**Author:** [@nirth\_AMD](https://devcommunity.amd.com/u/nirth_AMD)\
**Replies:** 0\
**Last updated:** [September 2, 2026, 9:49am UTC](https://devcommunity.amd.com/t/optimizing-atom-and-vllm-atom-for-high-interactivity-inference/944 "2026-09-02T09:49:34Z")
When you type a question into a chat assistant, ask a coding agent to make an edit, or watch tokens stream back from a model, what you feel is not throughput. You feel responsiveness. The pause before the first word appe…
---
## [LLM Inference at Scale on AMD Instinct: Multi-Node MoE, 64M-Token Context, and 4-bit KV Offloading](https://devcommunity.amd.com/t/llm-inference-at-scale-on-amd-instinct-multi-node-moe-64m-token-context-and-4-bit-kv-offloading/939)
**Author:** [@nirth\_AMD](https://devcommunity.amd.com/u/nirth_AMD)\
**Replies:** 0\
**Last updated:** [September 1, 2026, 4:11pm UTC](https://devcommunity.amd.com/t/llm-inference-at-scale-on-amd-instinct-multi-node-moe-64m-token-context-and-4-bit-kv-offloading/939 "2026-09-01T16:11:19Z")
Three recent AMD ROCm blogs tackle the same infrastructure problem from different angles: how do you actually serve large language models at scale when the KV cache becomes the bottleneck? Together they form a practical…
---
## [Serving NVFP4 Models on AMD Instinct™ MI355 Accelerators](https://devcommunity.amd.com/t/serving-nvfp4-models-on-amd-instinct-mi355-accelerators/726)
**Author:** [@nirth\_AMD](https://devcommunity.amd.com/u/nirth_AMD)\
**Replies:** 0\
**Last updated:** [July 20, 2026, 12:30pm UTC](https://devcommunity.amd.com/t/serving-nvfp4-models-on-amd-instinct-mi355-accelerators/726 "2026-07-20T12:30:39Z")
NVFP4 is an increasingly common deployment format: NVIDIA, AMD, and the open-source community have published NVFP4 quantized checkpoints of frontier models such as moonshotai/Kimi-K2.6, and many users want to deploy t…
---
## [Engineered a full-stack e-commerce application with AI-powered product description generation](https://devcommunity.amd.com/t/engineered-a-full-stack-e-commerce-application-with-ai-powered-product-description-generation/495)
**Author:** [@Sameer\_Kumar](https://devcommunity.amd.com/u/Sameer_Kumar)\
**Replies:** 1\
**Last updated:** [May 1, 2026, 12:55pm UTC](https://devcommunity.amd.com/t/engineered-a-full-stack-e-commerce-application-with-ai-powered-product-description-generation/495 "2026-05-01T12:55:54Z")
Built admin dashboard with dynamic product CRUD operations and AI-assisted content automation. ∗ Integrated Groq LLaMA 3.1 model via REST API for dynamic content creation. ∗ Developed optimized filtering, sorting, and re…
---
## [ROCm Becomes a First-Class Platform in the vLLM Ecosystem](https://devcommunity.amd.com/t/rocm-becomes-a-first-class-platform-in-the-vllm-ecosystem/321)
**Author:** [@Uncy\_AMD](https://devcommunity.amd.com/u/Uncy_AMD)\
**Replies:** 1\
**Last updated:** [March 11, 2026, 6:34pm UTC](https://devcommunity.amd.com/t/rocm-becomes-a-first-class-platform-in-the-vllm-ecosystem/321 "2026-03-11T18:34:19Z")
As the generative AI ecosystem matures, vLLM embraces a multivendor ecosystem. The quality of support across hardware platforms becomes a defining priority: developers expect consistent, high-performance behavior no m…