On May 27th, Lenovo Group's stock price rose by 8.86%, closing at 12.04 Hong Kong dollars per share, reaching a multi-year high, and even touched a high point of 12.26 Hong Kong dollars per share during the trading session. Last week, after Lenovo Group announced its full-year and fourth-quarter results for the fiscal year 23/24, Chairman and CEO Yang Yuanqing stated, "The most difficult period for the IT industry has passed," and in a letter to supply chain partners, he mentioned that artificial intelligence (AI) brings unprecedented development opportunities to the industry.

Lenovo previously put forward the "All for AI" strategy, which not only refers to seizing AI opportunities in products but also includes the application of AI technology in supply chain operations. Reporters from First Financial Daily learned that Lenovo has over 2000 core global suppliers, and the operation of this vast supply chain is being integrated with AI technology for the purpose of improving supply chain efficiency.

Xu He, Lenovo Group's Global Supply Chain Chief Transformation Officer and CEO of Liansheng Zhida, told First Financial Daily reporters that AI can improve the efficiency of supply chain operations, and artificial intelligence will also enter into production, distribution, and other processes in the future.

Advertisement

While AI brings opportunities, the supply chain also faces fluctuations. On the production side, a shortage of AI processors previously affected the shipment of some manufacturers. On the supply chain operation side, there are still some challenges for AI to be smoothly implemented in supply chain operations.

Landing AI in the Supply Chain

"The opportunities brought by AI should be limitless. It's not good to be slow (for companies). First, we must apply AI well, become the most intelligent enterprises, better predict demand, match supply and demand, make production scheduling more efficient, and make quality inspection and transportation more rational. This is what we need to do, and after doing it well, we can serve manufacturing enterprises," Yang Yuanqing said at a media communication meeting to media including First Financial Daily. The flexibility of the supply chain is one of the reasons for Lenovo's sustainable innovation, and the linkage between selling more and supplying more, as well as the intelligent demand forecasting, is related.

Lenovo's global supply chain covers the fields of PCs, servers, and mobile phones, with a total of over 4000 core suppliers and third and fourth-tier suppliers, and has more than 30 manufacturing bases worldwide. At the recent Lenovo Group Global Suppliers Conference, Du Zhemin, Senior Vice President and Group Operations Officer of Lenovo Group, introduced that before this round of AI boom, Lenovo Group has integrated AI into the global supply chain operations. The AI-driven Global Supply Chain Intelligent Control Tower (SCI) can reduce decision-making time by nearly 60%.

"Lenovo has brought the most core 400 suppliers to the digital collaboration platform, and through APIs (interfaces), big data analysis, and AI data mining technology, we summarize, share, and refine information, reducing fluctuations and bullwhip effects in supply chain management," Xu He said.

Xu He also told reporters that AI can improve work efficiency, quality, and stability at the levels of supply chain intelligent control towers and data analysis, providing early warnings. In the future, generative AI is expected to think and learn like humans, reduce human heavy physical labor, and enter into production, distribution, and R&D processes.

Xu He observed that there are differences in the motivation to apply AI and other advanced technologies at home and abroad. Domestically, the focus on digital transformation is more on the input-output ratio, considering opportunities at the cost level. Overseas, the application of AI or other advanced technologies is often due to labor restrictions, sometimes having to choose new technologies out of necessity.Xu He also pointed out that there are challenges for AI to be deeply applied in the supply chain. On the one hand, data involves privacy and security issues, and creating large models suitable for supply chain scenarios requires abundant data for training. Obtaining data in a reasonable and legal manner will be a significant challenge. On the other hand, relevant talent remains relatively scarce. In addition, Lenovo also needs to consider the relationship between AI and corporate social responsibility. Lenovo has previously driven employment in surrounding industries as the leader of the chain, and in the future, it will need to find a balance between technological iteration and the upgrading of labor capabilities.

Fluctuations in AI Server Business

As a server manufacturer, affected by AI demand, Lenovo's business is also fluctuating. In Lenovo's fiscal year 23/24, the Infrastructure Solutions Group (ISG), which includes the server business, suffered a loss for the year. Lenovo Group explained that changes in global IT budgets, along with the explosive growth of large models, have led to a rapid shift in cloud service providers' and enterprises' demands to AI-related computing, resulting in a decrease in traditional computing budgets, and a shortage of AI processor supply has also emerged.

Some server manufacturers have also been affected by the shortage of AI processor supply. In the first half of last year, Inspur Information's revenue decreased by 28.85% year-on-year, and net profit decreased by 65.91%. The financial report mentioned that the decline in revenue was mainly due to the tight supply of specialized chips and other factors. For the whole of last year, Inspur Information's revenue and net profit decreased by 5.41% and 14.54% year-on-year, respectively.

In addition, in June last year, H3C, a subsidiary of Unigroup, released AI servers and other products using NVIDIA H800 GPUs. However, the supply of related chips was insufficient afterward. In the first half of last year, H3C's revenue and net profit increased by 4.87% and 2.62% year-on-year, respectively, and the annual net profit growth narrowed to 4.27%, while net profit decreased by 8.58% year-on-year.

However, Lenovo Group explained that there are signs that the supply gap for AI processors has narrowed. Driven by the demand for AI servers, including Lenovo, the performance of several server manufacturers has also shown signs of recovery recently.

Lenovo's Infrastructure Solutions Group (ISG) saw a sales rebound in the fourth quarter of the fiscal year, with revenue increasing by 15.1% year-on-year. Liu Jun, Executive Vice President of Lenovo Group and President of the China region, explained that AI servers grew by more than 20%. In the first quarter of this year, Inspur Information's performance also recovered, with revenue and net profit increasing by 85.32% and 64.39% year-on-year, respectively. Foxconn's revenue decreased year-on-year for the whole of last year, but its revenue recovered growth in the first quarter of this year, with the revenue from generative AI servers increasing nearly three times year-on-year in the first quarter.

A server upstream supplier told the reporter that, according to the market situation they have encountered, the demand for overseas AI servers has grown more rapidly this year. There is also growth in the domestic market, but it should maintain a growth rate roughly the same as last year. This may be related to domestic internet companies being constrained by budgets. This year, the additional server demands from domestic customers are mostly related to AI servers, while the additional demand for traditional servers is relatively small. Some server upstream suppliers also told the reporter that the current cost of AI computing is relatively high, which affects the widespread application of AI. AI has only brought a part of the additional demand, and the strength of the server in driving the growth of related components may be limited.

TrendForce data also shows that global server shipments decreased by 6% year-on-year last year, and it is estimated to grow by 2.05% this year, mainly driven by U.S. cloud service companies. This year, the proportion of AI servers in the total server shipments is about 12.1%. Among the global demand for high-end AI servers by major cloud service companies this year, Microsoft, Google, Amazon AWS, and Meta account for more than 60% of the demand.