惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

博客园 - 【当耐特】
Threat Intelligence Blog | Flashpoint
Threat Intelligence Blog | Flashpoint
Y
Y Combinator Blog
D
DataBreaches.Net
Google DeepMind News
Google DeepMind News
H
Hackread – Cybersecurity News, Data Breaches, AI and More
云风的 BLOG
云风的 BLOG
Recorded Future
Recorded Future
I
InfoQ
L
LangChain Blog
Stack Overflow Blog
Stack Overflow Blog
Recent Announcements
Recent Announcements
宝玉的分享
宝玉的分享
Martin Fowler
Martin Fowler
J
Java Code Geeks
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
CTFtime.org: upcoming CTF events
CTFtime.org: upcoming CTF events
A
About on SuperTechFans
人人都是产品经理
人人都是产品经理
G
Google Developers Blog
大猫的无限游戏
大猫的无限游戏
C
Cybersecurity and Infrastructure Security Agency CISA
Know Your Adversary
Know Your Adversary
MongoDB | Blog
MongoDB | Blog
T
Tor Project blog
The Register - Security
The Register - Security
H
Help Net Security
Cisco Talos Blog
Cisco Talos Blog
P
Privacy & Cybersecurity Law Blog
NISL@THU
NISL@THU
P
Palo Alto Networks Blog
B
Blog RSS Feed
Latest news
Latest news
T
Threat Research - Cisco Blogs
The Hacker News
The Hacker News
C
Cisco Blogs
P
Privacy International News Feed
T
The Exploit Database - CXSecurity.com
V
Vulnerabilities – Threatpost
S
Schneier on Security
P
Proofpoint News Feed
Schneier on Security
Schneier on Security
www.infosecurity-magazine.com
www.infosecurity-magazine.com
cs.AI updates on arXiv.org
cs.AI updates on arXiv.org
AI
AI
Google Online Security Blog
Google Online Security Blog
H
Hacker News: Front Page
N
News and Events Feed by Topic
W
WeLiveSecurity

OpenCV

Live: Revolutionary Drone STEM Kit from DroneBlocks OpenCV 5 Is Here: The Biggest Leap in Years for Computer Vision How P&G Uses AI to Understand Human Behavior The Holographic Future Is Here. See It at OSCCA. Tangram Vision and OpenCV Are Partnering to Fix Your Calibration Problems Glenn Jocher of Ultralytics (YOLO) Is Speaking at OSCCA The Founder of OpenCV Is Speaking at OSCCA, Don’t Miss It! Behind the Magic: Disney Research Imagineering’s Doug Fidaleo Comes to OSCCA When the Track Is Your Lab: Meet the Team Racing Without a Driver Attend The OpenCV-SID Conference On Computer Vision & AI This May 4th Preview The Embedded Vision Summit 2026 Conference On OpenCV Live Calling Roboticists & Vision Experts: Tackle Dexterous Manipulation and Win Big in the AI for Industry Challenge
OpenCV and AMD Announce Collaboration to Accelerate Computer Vision and Vision AI Workloads on AMD Hardware
Phil Nelson · 2026-06-05 · via OpenCV


AMD is now an OpenCV 5 Launch Partner and will become an OpenCV Gold Sponsors as part of collaboration focused on OpenCV 5 CPU and GPU acceleration

PALO ALTO, CA. – June 4th 2026

OpenCV, the world’s leading open-source computer vision library, today announced a new engineering collaboration with AMD to make AMD hardware a first-class platform for OpenCV and Vision AI workloads. AMD is now an OpenCV 5 Launch Partner and will become an OpenCV Gold Sponsor as part of the agreement.

The collaboration will focus on accelerating the various stages of AI inference pipelines including pre- and post-processing, and other computer vision operations.

As an OpenCV 5 Launch Partner, AMD will work with OpenCV on key acceleration initiatives for the upcoming OpenCV 5 generation, including CPU optimization, GPU acceleration, benchmarking infrastructure, continuous integration, and upstream contributions to the OpenCV mainline repository.

A central focus of the collaboration is OpenCV 5’s new non-CPU Hardware Abstraction Layer, or HAL. This vendor-pluggable architecture is designed to enable dynamically loadable acceleration backends and transparent CPU fallback. Through this work, AMD is positioned as the first hardware platform targeted for official GPU support in OpenCV 5 for fast Vision AI pre- and post-processing pipelines.

On the CPU side, the engagement includes hand-optimized AVX-512 kernels for AMD Ryzen™ AI Embedded P100 Series systems based on “Zen 5” architecture. These kernels will target high-impact OpenCV core and imgproc operations using capabilities such as VNNI, VBMI, FMA, CD, Compress-Store, and related AVX-512 sub-extensions. The optimized kernels will be registered as separate CPU dispatch paths and selected automatically at runtime when supported AMD hardware is detected, while existing generic implementations remain available as fallbacks.

On the GPU side, OpenCV and AMD will work toward a HIP-based backend built on the AMD ROCm™ open software stack. The backend is intended to support AMD RDNA™ 3.5 integrated GPUs and RDNA 4 discrete GPUs, with HIP kernels, AMD-oriented runtime dispatch, zero-copy memory paths on unified-memory systems, and rocprof-based performance tuning. The work also creates a vendor-controlled acceleration path that can be developed, validated, and shipped independently while remaining aligned with OpenCV’s open-source architecture.

“At AMD, we’re focused on building open, high-performance software ecosystems that help developers bring the next generation of AI applications to life,” said KV Thanjavur Bhaaskar, Robotics Lead at AMD. “We’re excited to collaborate with OpenCV as an OpenCV 5 Launch Partner and Gold Sponsor to help accelerate computer vision and Vision AI workloads across AMD CPUs, integrated GPUs, and discrete GPUs. By working together on OpenCV 5’s new acceleration capabilities, we aim to make it easier for developers to build and deploy vision applications with the performance and efficiency they need on AMD platforms.”

The collaboration will include upstream contributions to OpenCV, performance benchmarking, CI/CD pipeline support, and validation on AMD hardware. The teams will also collaborate through a structured engineering process that includes regular pull request review, blocker resolution, resource triage for CI infrastructure, and feedback loops with the broader developer community.

“OpenCV 5 represents a major step forward for the computer vision community. As an OpenCV 5 Launch Partner, AMD will help bring high-performance CPU and GPU acceleration to the pre- and post-processing stages that power modern Vision AI systems,” said Satya Mallick, CEO of OpenCV. “We are excited to welcome AMD as an OpenCV Gold Sponsor and to work together on upstream contributions that strengthen OpenCV’s hardware neutrality while helping developers take full advantage of AMD platforms.”

By improving the OpenCV operations that prepare image and video data for AI models and process model outputs for downstream applications, the collaboration is expected to help reduce bottlenecks in end-to-end Vision AI systems. These improvements are relevant across robotics, industrial automation, smart cameras, embedded systems, edge AI, healthcare imaging, retail analytics, and other computer vision use cases.

OpenCV Gold Sponsorship will allow AMD to further support OpenCV’s open-source development, infrastructure, and community programs. Gold Sponsors play an important role in helping OpenCV maintain and advance the library used by developers, researchers, companies, startups, and academic institutions worldwide.

The collaboration reflects a shared commitment to open ecosystems, high-performance computing, and practical AI deployment. Together, OpenCV and AMD will work to strengthen the performance and reliability of computer vision workloads on AMD platforms while contributing improvements back to the broader OpenCV community.

About OpenCV

OpenCV is the world’s leading open-source computer vision library and one of the largest open-source computer vision frameworks, with more than 40 million monthly downloads. OpenCV provides developers, researchers, and organizations with tools for real-time image processing, computer vision, and AI-powered perception across robotics, industrial automation, healthcare, manufacturing, security, consumer devices, education, and research.

Media Contacts

AMD, the AMD arrow logo, RDNA, ROCm, Ryzen and combinations thereof are trademarks of Advanced Micro Devices, Inc.