Huawei says its delayed public roll-out of a cloud platform was a deliberate strategic move to ensure scalability, in-house R&D integration and the development of transformation expertise it now offers globally. The company believes this approach positions it well as the market shifts from cloud adoption to scale, and as convergence between cloud, AI and data centre ecosystems accelerates.
Speaking at the sixth annual Huawei ICT Editors Xchange, held in Johannesburg on 11 June, Diego Han, director of Ecosystem Development at Huawei Cloud Sub-Saharan Africa, said running its own public cloud was critical in building technical depth and transformation capabilities.
Han reaffirmed Huawei’s focus on secure, scalable cloud infrastructure, intelligent network integration and AI platforms underpinned by trust and interoperability.
Charles Cheng, deputy CEO of Huawei South Africa, echoed the company’s long-term commitment to building open, interconnected and scalable digital ecosystems.
“Digital ecosystems matter insofar as they enable a better, more fulfilling life,” Cheng said. “They must be interconnected, adaptive and built to scale. We are talking about technology that defines tomorrow’s digital ecosystem.”
Cheng said this requires high-performance computing and access to converged ecosystems, forming the backbone of Huawei’s "AI-native everything as a service" strategy. This also supports the company’s transition from cloud-native to AI-native services.
Pangu LLM and DeepSeek collaboration
In June 2024, Huawei Cloud confirmed its Pangu large language model (LLM) could be launched in SA, contingent on market demand. Calvin Huang, head of Solutions Architecture at Huawei Cloud SA, said the company is investing in AI-native infrastructure to support widespread AI adoption and meet rising computing demands.
This AI push is bolstered by a collaboration with DeepSeek, an open source AI model. In May, Huawei Cloud stated its end-to-end AI enablement stack is powered by DeepSeek and the DeepSeek R1 AI engine.
“AI is changing everything,” said Rex Lei, president of Huawei Cloud sub-Saharan Africa. “As an open source model, DeepSeek is growing rapidly, reaching nearly 100 million registered users in record time. It gives countries and companies the tools to develop their own models.”
Lei noted that Huawei invests over 20% of annual revenue in R&D, with a strong focus on cloud and AI. Huawei Cloud's strategy prioritises AI, positioning the company as an innovation leader in the transition to AI-native infrastructure.
Huawei Cloud defines this strategy through two lenses: "AI for cloud", where services are re-engineered using AI, and "cloud for AI", where Huawei builds foundational infrastructure – computing, data, tools and services – to accelerate AI development.
Local cloud expansion and internal transformation
Huawei was the first hyperscale provider to open a data centre region in SA, in 2019. The company says since then, its public cloud business has grown more than 16-fold, placing it among the top three local providers alongside Microsoft Azure and Google Cloud.
Han emphasised Huawei’s continued partnerships with local telecommunications providers, which have integrated Huawei’s infrastructure into customer-facing solutions. Internally, the company has used its cloud services to accelerate digital transformation.
Between 2015 and 2018, Huawei reduced the development cycle of its internal IT applications from six to nine months to between one week and one month, boosting overall service delivery efficiency by 30%, Han said.
AI in healthcare: Bridging the access divide
Tania Joffe, founder and principal of South African health tech platform Unu Health, demonstrated real-world use cases of AI at the event, emphasising its role in improving healthcare access in underserved communities.
“There is no business that will be untouched by the shifts taking place now,” said Joffe. Unu Health leverages AI to offer mobile diagnostic tools – such as facial scans that detect blood pressure, oxygen levels and pulse – as well as AI-assisted diagnostics and monitoring.
“South Africa has a two-tier healthcare system. AI can help bridge that divide – but success depends on more than algorithms. You need smartphones, connectivity and trust in the system,” she said.
Joffe compared the potential of digital healthcare to vehicle diagnostics, and said just as consumers want data-driven answers when servicing a car, the same applies to healthcare. She cited the use of LLMs as digital twins to better understand and respond to the unmet needs of target populations.
However, she stressed that infrastructure and responsible data practices must accompany innovation. “This is a moment of real possibility, but only if we’re intentional about building systems that protect data, empower people and deliver equitable care at scale.”
Share