Posts tagged Computer Vision
Ultralytics YOLO on AMD GPUs: Train, Run, and Deploy with ROCm™
- 26 October 2026
Ultralytics YOLO is the most popular open-source framework for real-time computer vision. With just a few lines of Python, developers can train and deploy YOLO models for a range of tasks, including object detection, segmentation, tracking, pose estimation, and image classification.
Accelerating MONAI Digital-Pathology Pipelines on AMD Instinct™ GPUs with hipCIM
- 26 October 2026
A digital-pathology model is only as fast as the pipeline that feeds it. That’s simple to overlook when the conversation turns to network architectures and FLOPs, but on real whole-slide workloads the truth is stubborn: a great deal of the wall-clock time goes to reading gigapixel slides off disk and decoding them into the small tiles the model consumes. Speed up the model and leave the reader untouched, and the pipeline barely moves.
Scaling 3D Gaussian Splatting with Distributed Compute: Why One GPU Is Never Enough
- 23 October 2026
gsplat is a GPU-optimized Python library for training and rendering 3D Gaussian Splatting (3DGS) models. In earlier posts we showed how to train gsplat scenes on AMD GPUs, took a closer look at the mathematics of 3DGS, and applied gsplat to autonomous driving.
Micro-World 2: Streaming Real-time Interactive World Model
- 21 October 2026
Diffusion models have advanced rapidly in recent years, enabling high-quality video generation across a wide range of creative scenarios. As visual quality continues to improve, however, the field is increasingly shifting its focus from passive content generation toward controllable and interactive experiences. YUME 1.5 [1] explores controllable video generation through explicit interaction signals, while LingBot-World [2] and Sana-WM [3] investigate world models that generate future observations conditioned on the current environment and user actions. World Labs’ Marble [27] and Atlas [28] represent a complementary spatial-modeling direction: Marble creates editable, explicit 3D worlds from multimodal inputs, while Atlas organizes text, images, camera poses, and depth into a shared spatial context for generation, reconstruction, and simulation. Together, these efforts illustrate how generative models are evolving from passive content-synthesis tools toward controllable, persistent, and interactive world simulators.
hipCIM 26.06: GPU-Accelerated Whole-Slide Image I/O on AMD Instinct™ GPUs
- 20 October 2026
Every whole-slide image tells a story in pixels. A whole-slide image is a glass pathology slide scanned into a single very large digital image, stored as a pyramid of small tiles so a viewer can pan and zoom it. A single biopsy digitized at 20× magnification can span tens of thousands of pixels on each side and weigh in at several gigabytes. Before a model can learn anything from that slide, something has to read it off disk, decode its thousands of small JPEG-compressed tiles, and turn them into tensors the accelerator can work on. In digital pathology, that unglamorous first step is often where the wall-clock time quietly disappears.
ROCm 10.1: Breaking the Data-Movement Bottleneck
- 05 October 2026
What limits your training run today: how fast the GPU computes, or how fast you can get data to it? For a growing number of workloads, the answer is the latter. ROCm 10.1 leads with that path, and carries a full slate of compiler, library, and tooling work behind it.
ROCm 10.0: A Decade of Open Compute, Built for the Age of Agentic AI
- 27 August 2026
AMD shipped ROCm 1.0 in April 2016: an open-source GPU compute stack built around a C++ compiler and a GPU programming language called HIP, aimed at high-performance computing. A decade later, the same platform trains and serves frontier AI models across industries and enables nearly every compute application benefiting from GPUs.
AUP Learning Cloud: Streamlining AI Education on AMD
- 06 August 2026
AI education is becoming increasingly hands-on. Students are now expected to train models, build AI agents, experiment with large language models (LLMs), and develop applications accelerated by graphics processing units (GPUs), so courses depend as much on practical computing infrastructure as on learning materials. Yet standing up that infrastructure is usually the real bottleneck, and educators often spend more effort building it than teaching.
Efficient Hyperparameter Optimization for Autonomous Driving Models with AMD Instinct GPU Partitioning
- 08 July 2026
For automotive OEMs and Tier-1 suppliers, developing production-grade perception models presents a significant bottleneck: achieving the accuracy required for safety-critical applications such as forward collision warning and autonomous emergency braking demands systematic hyperparameter optimization (HPO). However, HPO requires training hundreds of model variants, each taking hours on a single GPU, making thorough exploration of the parameter space prohibitively slow and expensive with conventional single-GPU setups.
Building a GPU-Resident YOLO26 Object Detection Pipeline on the AMD Radeon™ AI PRO R9700 GPU
- 03 July 2026
Modern AMD GPUs include a dedicated hardware block for video processing called the Video Core Next (VCN) engine. By chaining VCN directly into machine learning frameworks, you can build an object detection pipeline where a video frame stays in VRAM from decode to the final bounding boxes. The host sees only the surviving detections.
Semantic Fencing of Video Streams Using Embedding Splits from Vision Foundation Models
- 15 May 2026
In this blog, we present a novel approach for semantically splitting vision datasets into training, validation, and test sets. Instead of relying on ad hoc metadata rules or random shuffles, we use embeddings to reason directly about similarity in latent space and construct splits that better reflect true generalization.
3D Scene Reconstruction from the Inside: Explore the Mathematics Behind gsplat
- 16 December 2025
3D Gaussian Splatting (3DGS) reconstructs 3D scenes from multiple 2D images and renders novel views in real time. In this blog, which serves as a follow up to a previous post, Elevating 3D Scene Rendering with GSplat, you will learn the core mathematics and the practical library components behind 3DGS using gsplat.
Medical Imaging on MI300X: SwinUNETR Inference Optimization
- 10 December 2025
This post is the inference companion to our training walkthrough, Medical Imaging on MI300X: Optimized SwinUNETR for Tumor Detection. In this post, we focus on inference latency and throughput for 3D CT segmentation of tumors on MI300X, and we share the two optimizations that delivered the biggest wins in practice:
STX-B0T: Real-time AI Robot Assistant Powered by RyzenAI and ROCm
- 23 October 2025
In this blog post, we introduce STX-B0T, our human-responsive social robot prototype powered by AMD Ryzen AI’s CVMLSDK, the Strix-Halo (STX-Halo) APU, and built with open-source hardware.
Medical Imaging on MI300X: Optimized SwinUNETR for Tumor Detection
- 07 October 2025
This blog is part of a series of walkthroughs of Life Science AI models, stemming from this article.
Announcing MONAI 1.0.0 for AMD ROCm: Breakthrough AI Acceleration for Medical Imaging Models on AMD Instinct™ GPUs
- 07 October 2025
Today, AMD is thrilled to announce MONAI 1.0.0 for AMD ROCm, now available to the community as part of the ROCm-LS Early Access release. In this blog you will learn how to use MONAI to load an analyze medical images, and see results from more advanced case studies.
Elevating 3D Scene Rendering with GSplat
- 03 October 2025
In this blog we explore how to use GSplat, a GPU-optimized Python library for training and rendering 3DGS models, on AMD devices. This tutorial will guide you through training a model of a scene from a set of captured images, which will then allow you to render novel views of the scene. We use a port of the original GSplat code that has been optimized for AMD GPUs. The examples used throughout this blog were trained and rendered using an AMD MI300X GPU.
ROCm 7.0: An AI-Ready Powerhouse for Performance, Efficiency, and Productivity
- 16 September 2025
Artificial intelligence now defines the performance envelope for modern computation. In this blog, we introduce the AI-centric ROCm 7.0 designed to help our community directly benefit from this dramatic paradigm shift. ROCm 7.0 delivers a platform purpose-built for the era of generative AI, large-scale inference and training, and accelerated discovery, helping you boost the performance, efficiency, and scalability of your workloads.
Announcing hipCIM: A Cutting-Edge Solution for Accelerated Multidimensional Image Processing
- 18 July 2025
In the rapidly evolving landscape of data science and computational imaging, hipCIM 1.0.0 introduces a powerful, GPU-accelerated open-source library that redefines multidimensional image processing for life sciences, biomedical research, and computational imaging. This open-source, accelerated software library redefines how multidimensional datasets are processed, offering unparalleled capabilities across scientific fields such as biomedical imaging, geospatial analytics, material sciences, life sciences, and remote sensing to name a few. With the initial release of hipCIM 1.0.0, AMD enters the arena, ready to push the boundaries of life science research and stand at the forefront of a new era in multidimensional image processing.
Accelerated JPEG decoding on AMD Instinct™ GPUs with rocJPEG
- 12 May 2025
With the increased growth in dataset sizes, the improvement of image capturing technology, the capacity to extract more information from visual data, and the move towards large language models including image data as input, efficient image processing and preparation has become a necessity to run these workloads in a timely manner. Although much attention is often focused on the computational aspects of these workloads, the fundamental tasks of data loading and preparation have become significant bottlenecks, limiting the throughput of the entire pipeline. Accelerated JPEG decoding is an essential step in optimizing workloads that rely on image data. Dive into this blog post to learn how to install and benchmark rocJPEG, as well as how the ROCm™ platform and AMD Instinct GPUs can help you achieve up to 50x faster decoding performance in 4k1.
ROCm 6.4: Breaking Barriers in AI, HPC, and Modular GPU Software
- 11 April 2025
In the rapidly evolving landscape of high-performance computing and artificial intelligence, innovation is the currency of progress. AMD’s ROCm 6.4 isn’t just another software update—it’s a leap forward that redefines the boundaries of what is possible for AI, developers, researchers, and enterprise innovators.
Vision Mamba on AMD GPU with ROCm
- 24 January 2025
State Space Models (SSMs), such as Mamba, have emerged as a potential alternative to Transformer models. Vision backbones using only SSMs have yielded promising results. For more information about SSMs and Mamba’s performance on AMD hardware, see Mamba on AMD GPUs with ROCm. This blog explores Vision Mamba (Vim), an innovative and efficient backbone for vision tasks and evaluate its performance on AMD GPUs with ROCm. We’ll start with a brief introduction to Vision Mamba, followed by a step-by-step guide on training and running inference with Vision Mamba on AMD GPUs using ROCm.
Image Classification with BEiT, MobileNet, and EfficientNet using ROCm on AMD GPUs
- 03 September 2024
Image classification is a key task in computer vision aiming at “understanding” an entire image. The outcome of an image classifier is a label or a category for the image as a whole, unlike object recognition where the task is to detect and classify multiple objects within an image.
Accelerate PyTorch Models using torch.compile on AMD GPUs with ROCm
- 11 July 2024
PyTorch 2.0 introduces torch.compile(), a tool to vastly accelerate PyTorch code and models. By converting PyTorch code into highly optimized kernels, torch.compile delivers substantial performance improvements with minimal changes to the existing codebase. This feature allows for precise optimization of individual functions, entire modules, and complex training loops, providing a versatile and powerful tool for enhancing computational efficiency.
Panoptic segmentation and instance segmentation with Detectron2 on AMD GPUs
- 23 May 2024
23, May 2024 by .
Unlocking Vision-Text Dual-Encoding: Multi-GPU Training of a CLIP-Like Model
- 24 April 2024
24 Apr, 2024 by .
Interacting with Contrastive Language-Image Pre-Training (CLIP) model on AMD GPU
- 16 April 2024
16, Apr 2024 by .