Main content

KubeCon + CloudNativeCon China – news and events round-up

Chris Middleton Profile picture for user cmiddleton June 12, 2025
Summary:
We present a concise round-up of news and announcements from our in-depth access to this major event in Hong Kong

an image of the Hong Kong skyline at night, surrounded by clouds

KubeCon + CloudNativeCon China 2025 has hosted two days of keynotes, presentations, and lightning talks, with the main and exhibition halls and the breakout rooms at The Hopewell bustling with conversation and enthusiasm.

On day one, Bill Ren (Ren Xudong), Chief Open Source Liaison Officer for Chinese electronics giant, Huawei, compared AI with the arrival of steam power in the 19th Century. But he adds a caveat, saying:

But it took another 70 years until what we call the Industrial Era arrived, with real applications for the steam engine.

The same principle applied to Faraday's discovery of magnetic induction in 1831, he says: it was a half century before the planet, industry, and human lives began being transformed by electricity. The lesson? The real benefits of these early days of the AI Spring may be hard to predict or take years to arrive – even at internet speed.

But the open source movement will be at the center of that transformation when it comes, he predicts, enabling the underlying infrastructure and making it more efficient – helping it arrive sooner, in fact, via clouds of AI clusters.

On that point, the Linux Foundation used KubeCon + CloudNativeCon China to stress the findings of its May 2025 report on open source AI (OSAI) and its predicted impact on economies and jobs. Headline findings are that 89% of all organizations are using some form of open source in their AI stack, with nearly two-thirds using an open model.

The technical advantages lie in cost effectiveness, productivity, and "accelerating collaborative innovation", claims the report, with "nuanced" complementary impacts on employment.

Like many vendors, the report claims "unique" impacts of AI on healthcare, energy, manufacturing, and other sectors, yet it is hard to avoid the impression that the technology is being sold on the back of industrializing creative processes.

That strategy seems likely to damage uptake of AI in the short term, especially if the 50+ lawsuits for copyright theft find for the plaintiffs. The industry would do well to recognize these issues rather than ignore them.

Building Asia's cloud native ecosystem

Other Chinese giants presented on Day One, including Alibaba Cloud and TikTok parent, ByteDance, which is managing large-scale LLM inference workload on Kubernetes. PayPal, Akamai, DaoCloud, Microsoft, and KPMG all hosted technical Lightning Talks.

Meanwhile, Horizon Robotics explained how it is using open source for end-to-end, high-performance training in autonomous driving technology – via over 2,000 GPUs – and voice recognition specialist iFlyTek, partly owned by the Chinese government, is training AI models at massive scale, using Volcano to cut failure recovery time by 70%.

On Day Two, Bilibili, Cisco, Google, Bloomberg, China Mobile, Intel, and NVIDIA were among the heavyweights exploring the technical aspects of open source cloud integration, AI, and more.

KubeCon certainly feels welcoming to allcomers, wherever the conference takes place in the world – next stop is Tokyo, Japan, on 16-17 June (but not for diginomica, alas). The sense that this is a real community, both in China and in the open source world outside, is palpable.

The Cloud Native Computing Foundation (CNCF) annual report certainly finds the organization going from strength to strength, with the first KubeCon in India, plus the launch of the 'Kubestronaut' program (and Kubernetes itself turning ten – or is that #KuberTENes?)

Eighty percent of organizations plan to build most new applications on cloud-native platforms within five years, according to CNCF research. Yet three-quarters report challenges in Kubernetes adoption due to a lack of skilled talent. The Kubestronaut program helps address this need by "recognizing individuals who validate their skills through CNCF certification", says the organization.

CNCF CTO Chris Aniszczyk explains:

The Kubestronaut program is about recognizing commitment and the passion these individuals demonstrate as they build their knowledge in the cloud native space. In Asia especially, we're seeing developers turning certification into a springboard for leadership and deeper engagement in the cloud-native ecosystem."

Since launching in November 2024, CNCF has recognized nearly 1,800 Kubestronauts across over 100 countries. Asia accounts for 280 of that total, including 29 in China and 63 in Japan.

Together, mainland China and Hong Kong have been one of CNCF's strongest ecosystems, explains Keith Chan, the Linux Foundation's Director of Strategic Planning. It continues to rank second in the world, with 1,068,553 code contributions, though a long way behind the US' 5.3 million.

One landmark project has been China's launch of a space computing satellite, using open source provider CubeFS to manage data distributed across satellite constellations. China's ambitions, it seems, are not earthbound.

My take

So long from Hong Kong, though we will return to some of the case studies and conference findings in the weeks ahead. For me, the next priority is avoiding the typhoon that is about to make landfall. Wish me luck!

Image credit - Image by carloyuen from Pixabay

Loading
A grey colored placeholder image