DynamiaDynamia
Product

Dynamia HAMi x Kunlunxin P800 Complete Compatibility Mutual Recognition

DynamiaProduct
Dynamia HAMi x Kunlunxin P800 Complete Compatibility Mutual RecognitionProduct capabilities and release notes
Dynamia · Open-source AI infrastructure

Recently, Dynamia and Kunlunxin (XPU) completed phased product compatibility testing and mutual recognition around collaborative adaptation and engineering validation of domestic AI accelerators and AI infrastructure software.

On this page

Product Compatibility Mutual Recognition

Recently, Shanghai Dynamia Intelligent Technology Co., Ltd. (hereinafter "Dynamia") and Kunlunxin (Beijing) Technology Co., Ltd. (hereinafter "Kunlunxin") completed phased product compatibility testing and mutual recognition around collaborative adaptation and engineering validation of domestic AI accelerators and AI infrastructure software.

Through joint testing and validation by both teams, Dynamia's HAMi has completed compatibility adaptation with the Kunlunxin P800. It performed well across related test items including functional availability, system stability, resource scheduling, and compatibility, and can support typical production scenarios such as AI inference and model training.

This compatibility mutual recognition was completed through collaboration in real engineering environments. During adaptation and validation, the Kunlunxin team actively participated in HAMi's technical integration and test verification, conducting joint tuning around key capabilities such as heterogeneous compute scheduling, resource isolation, and efficiency improvement, and advancing the practical adoptability and continuous evolution of these capabilities in real business scenarios.

Going forward, Dynamia will continue to advance HAMi's adaptation and mutual certification with more domestic AI chip and accelerator vendors. By combining open-source community and industry collaboration, it aims to provide users with stable, open, and selectable domestic compute infrastructure solutions and help AI applications scale into production.

Product compatibility certification certificate
Figure 1: Product compatibility certification certificate

Enterprise Products Also Complete Kunlunxin Adaptation

At the same time, based on the HAMi technology foundation, Dynamia's HAMi Enterprise Edition and HAMi Platform Edition have also completed adaptation and validation for the Kunlunxin P800, and can run stably in real production environments.

While maintaining consistency with HAMi's open-source capabilities, the enterprise products further strengthen stability, operations management, and production assurance for large-scale deployment, supporting enterprises and intelligent computing centers in meeting end-to-end needs—from PoC validation to production rollout—when using domestic AI accelerators such as Kunlunxin.

About Kunlunxin

Kunlunxin (Beijing) Technology Co., Ltd. originated from Baidu's Intelligent Chip and Architecture Department. It completed an independent financing round in April 2021 at a first-round valuation of about RMB 13 billion. The team was among the earliest in China to enter AI acceleration, with more than a decade of deep work, and is an AI chip company with strong accumulation across architecture, chip implementation, software systems, and scenario applications.

About Dynamia

Shanghai Dynamia Intelligent Technology Co., Ltd. focuses on GPU virtualization and heterogeneous compute scheduling to improve compute utilization in AI scenarios. The company initiated and leads the CNCF open-source project HAMi—the industry's only open-source project that delivers flexible, on-demand, elastic, and reliable GPU virtualization—and already supports mainstream AI chip ecosystems.

KEEP READING

Keep reading

All articles
DynamiaEngineering
As Models Grow Larger, Why Do Enterprises Still Need Fine-Grained Compute Scheduling?Deep dives and engineering detail
Dynamia · Open-source AI infrastructure

As Models Grow Larger, Why Do Enterprises Still Need Fine-Grained Compute Scheduling?

In a shared resource pool, enterprises must support full-GPU and multi-GPU deployments for large models while also enabling on-demand sharing for lightweight inference, R&D, and batch jobs. Drawing on model usage data and production cluster research, this article discusses how to allocate GPUs by workload needs and manage heterogeneous compute shared across teams.

Read article
DynamiaCommunity
Technical Progress and Hands-on Practice: HAMi at KubeCon NA 2026Building together · discussions · event notes
Dynamia · Open-source AI infrastructure

Technical Progress and Hands-on Practice: HAMi at KubeCon NA 2026

At KubeCon + CloudNativeCon North America 2026, the HAMi community joined Red Hat to discuss GPU sharing and inference architecture, covered scheduling policies and Kubernetes DRA progress, and led developers through GPU virtualization in ContribFest.

Read article
DynamiaCommunity
Making the Most of Every GPU: Dynamia Shares Practices at ContainerDays SingaporeBuilding together · discussions · event notes
Dynamia · Open-source AI infrastructure

Making the Most of Every GPU: Dynamia Shares Practices at ContainerDays Singapore

At ContainerDays & AI Context Singapore 2026, Dynamia shared HAMi GPU sharing and scheduling practices in Kubernetes through hands-on labs and production cases, plus GPU utilization optimization for Neocloud scenarios.

Read article
Image viewer