Posts by Jiangyong Ren
Production-Ready MXFP4 Online Rotation with Fused Kernels on AMD Instinct™ MI355X
- 13 August 2026
Serving large language models affordably increasingly depends on low-bit quantization, and MXFP4 is one of the most aggressive options — but the smaller models that need it most rely on online rotation to stay accurate, and that rotation has historically carried a steep latency tax. In this post you will learn how a single fused Gluon (Triton) kernel on AMD Instinct™ MI355X (CDNA4) removes that tax: we walk through the kernel-fusion design, the RS=64 optimization, the GEAK + Hyperloom tuning workflow, and end-to-end measurements showing online-rotation overhead falling from a prohibitive +5–10% to just +0.3–0.8% on Dense models — with no measurable accuracy loss. By the end, you will understand how to make accuracy-preserving MXFP4 online rotation cheap enough to ship by default.
VSA: Accelerating Video Diffusion Inference with Sparse Attention on AMD GPUs
- 04 August 2026
Video generation powered by diffusion transformers has achieved remarkable quality, but the computational cost of attention mechanisms remains a critical bottleneck. With sequence lengths reaching tens of thousands of tokens in video generation tasks, the quadratic complexity of standard attention becomes prohibitively expensive.
Advanced MXFP4 Quantization: Combining Fine-Tuned Rotations with SmoothQuant for Near-Lossless Compression
- 17 February 2026
As language models continue to grow in popularity, reducing the cost of inference and accelerating model serving have become key challenges. Quantization offers a powerful solution by reducing the model size and leveraging inexpensive math operations, for example, using low-bitwidth formats like OCP MXFP4 (4.25 bits) available in AMD Instinct MI350X and MI355X accelerators.
QuickReduce: Up to 3x Faster All-reduce for vLLM and SGLang
- 26 August 2025
Advancements in large-scale language models (LLMs) have led to significant performance breakthroughs across various domains, especially in natural language processing. LLMs typically consist of billions of parameters, resulting in substantial computational, storage, and deployment challenges. Inter-GPU communication overhead often emerges as a key bottleneck limiting overall system performance. In tensor-parallel setups, every layer requires frequent all-reduce operations—synchronizing large amounts of data across GPUs. This introduces significant latency and strains interconnect bandwidth.