Several Chinese cloud providers have hiked up the prices for their cloud and AI services.

As reported by The Register, Alibaba Cloud has increased prices for many services by up to 34 percent, while both Tencent and Baidu have followed suit with their own increases.

Alibaba Cloud was the first to make the move, sharing a price adjustment notice on March 18 for its "AI computing power and storage products." The increases vary, with GPU-based services increasing between five and 34 percent, and the company's CPFS (Intelligent Computing Edition) services increasing by 30 percent across the board.

These increases are impacting both third-party chips, and Alibaba's home-created AI chips.

The notice puts the increase down to a "surge in global demand and rising supply chain costs," which have meant "the procurement costs of core hardware in the industry have increased significantly."

Just one day later, Baidu Cloud followed suit with its own statement that reads: "Driven by the rapid development of global artificial intelligence applications, the demand for computing power continues to rise. The cost of core hardware and related infrastructure has increased significantly. To ensure the long-term stable operation of the platform and service quality, we have structurally optimized the pricing of some products."

The company has similarly increased prices by between five and 30 percent for AI computing power-related products and services, and 30 percent for parallel file storage.

Tencent Cloud, meanwhile, saw president Martin Lau speaking on the matter in its earnings call on March 18.

Lau said that the suppliers of data center equipment are prioritizing larger companies - hyperscalers, including Tencent - and "therefore … smaller cloud providers no longer have certainty that they can source supply, and they need to come to the hyperscalers. You know, the hyperscalers have been operating at low margins and so, you know, when the demand picks up, then, you know, we almost sort of as an industry have no choice but to pass through higher prices.”

Lau noted that this applies more for training chips, and as the industry moves towards inferencing, they will be able to earn at a lower margin.

The Chinese cloud providers are not alone in increasing prices - at the start of the year, Amazon Web Services (AWS) increased prices for some of its instances based on the Nvidia H200 GPU. At the time, the company said: "EC2 Capacity Blocks for ML pricing are dynamic and vary based on supply and demand patterns, as described on the product detail page. This price adjustment reflects the supply/demand patterns we expect this quarter."

The shift in China's cloud computing pricing, however, does recall to memory a similar dramatic change of a couple of years ago, if in a different direction.

Almost exactly two years ago, Alibaba Cloud dramatically cut its cloud computing costs for more than 100 products, in some cases by as much as 55 percent. Similarly, Tencent, JD.com, and Baidu had significant cuts to services. These reductions were not solely linked to AI workloads, however.

This was a stunt by the companies to gain customers. At the time, JD posted to WeChat to declare: "Cut all you like, let’s fight to the end!”