For the first time, Qwen3.8 brings a Qwen-Max-class model to open release. Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Beyond answering harder questions, Qwen3.8 is designed to carry complex, multi-step tasks through to completion with greater reliability. Qwen3.8 features the following enhancements: Core Capabilities: Comprehensive improvements across coding, professional work, research, and long-horizon agentic tasks. Agent Execution: Stronger autonomous planning and better ha
Day 0 Support for Qwen 3.8 on AMD Instinct GPUsZenDNN 5.2.1 builds on the modular multi-backend architecture introduced in 5.2, extending it with deeper quantization paths and kernel-level optimizations across the ZenDNN runtime.
ZenDNN 5.2.1: Deepening Quantization and Expanding the AI Inference Frontier on AMD EPYC™ CPUsMiniMax M3 is a new open-weight model for coding, agentic, and multimodal workloads. MiniMax describes M3 as combining three frontier capabilities in one model: strong coding and agentic task performance, long-context MiniMax Sparse Attention (MSA), and native multimodality for text, image, and video understanding. The 1M-token context window of MiniMax M3 enables sophisticated, long-horizon application workflows, including autonomous software engineering agents, repository-wide reasoning, and native multimodal document analysis alongside tool-driven automation. The AMD day-zero enablement foc
Day 0 Support for MiniMax M3 on AMD Instinct GPUs… With ZenDNN 5.2.1, this is implemented through a complete TorchAO-based pipeline using Int4WeightOnlyOpaqueTensorConfig , including: Full asymmetric 4-bit weight representation with zero-point handling Bias support for asymmetric quantized operations Operator fusion for quantized linear layers, hel… …
… TP=8 + fp8 KV is the practical combo at 1M: SGLANG USE AITER=1 sglang serve \ --model-path MiniMaxAI/MiniMax-M3 \ --quantization mxfp8 --kv-cache-dtype fp8 e4m3 --context-length 1048576 \ --attention-backend triton --moe-runner-backend triton \ --page-size 128 --disable-radix-cache --chunked-prefil… …
… By following this guide, you have learned how to deploy MiMo-V2.5-Pro using SGLang/ATOM to leverage advanced tool-calling capabilities for agentic workloads, as well as how to run coding and Skills-driven applications locally with this flagship model. …
… August 06, 2026 Quark Support for HuggingFace Diffusers and SVDQuant — ROCm Blogs Learn how to quantize, save, and reload diffusion models in Quark using its new SVDQuant and HuggingFace Diffusers support. …
… The MXFP4 checkpoint can be reproduced with the following Quark quantization flow: from quark.torch import LLMTemplate, ModelQuantizer qwen3 5 moe text template = LLMTemplate model type="qwen3 5 moe text", kv layers name= " k proj", " v proj" , q layer name=" q proj" LLMTemplate.register template q… …
… The entry point in quantum computers is called the Quantum Controller, a piece of hardware that bridges the gap between the digital world of traditional computers and the analog world of quantum. Most quantum computers nowadays implement the Quantum Control unit via AMD FPGAs. …
… Product Marketing Manager, AI Group Contributors Jereshea John Mary Arijit Mukhopadhyay Fellow, System Design Engineering Avinash-Chandra Pandey Chandra Kumar Ramasamy Naveen Thangudu Yogesh Vidhya Muthu Related Blogs View All Blogs Larger-Than-Ever Single-GPU Quantum Simulation: How AMD and Qibo A… …
… August 06, 2026 Quark Support for HuggingFace Diffusers and SVDQuant — ROCm Blogs Learn how to quantize, save, and reload diffusion models in Quark using its new SVDQuant and HuggingFace Diffusers support. …
… May 19, 2026 Larger-Than-Ever Single-GPU Quantum Simulation: How AMD and Qibo Are Pushing the Boundaries of Quantum Simulation and Control In this blog, we highlight our ongoing synergy with Qibo, an end-to-end open-source framework for quantum computing, and In this blog post, we present the large… …
… Figure 2: InferenceX Performance pareto curve - throughput vs interactivity for AMD Instinct™ MI355X MoRI SGLang vs B200 Dynamo SGLang What You'll Learn Under the Hood How AITER FlyDSL empowers MoE in distributed scenarios How MoRI quantized dispatch/combine benefits expert parallelism How MoRI KV … …