Computing supply chain firms respond to AI demand slowdown rumors: no cuts and price hikes seen

On November 4, 2026 at 08:32, Eastmoney reported on how several compute-industry companies are pushing back against market rumors that major tech companies are slowing their AI spending. The companies say they see no signs of a pullback and describe their order books as solid. Xiechuang Data stated that it currently sees no downward adjustment, and expects capital expenditure at leading Chinese internet firms to remain at a high level next year, keeping compute infrastructure investment robust. On the optical module side, the company said it is jointly developing super-node servers with partners across the server and GPU supply chain, and that the project is progressing smoothly. Xingyun Technology said that the compute and storage orders it has publicly disclosed this year have reached approximately CNY 20.641 billion. Its top customers — referred to as Customer V, Customer VB and Customer VC — have negotiated higher lease prices during project execution in recognition of the company's delivery capability, with Customer V and Customer VB having already completed a second contract price increase. GigaDevice attributed the strength to surging demand for memory chips driven by rapid AI development, along with spillover effects. It noted that major manufacturers are cutting or shifting capacity and exiting niche markets, so supply of niche memory products will remain tight. For NOR Flash, it expects prices of high-capacity products to keep rising modestly for a period, while prices of small and mid-capacity products stay stable. For SLC NAND Flash, the long lead time for capacity to come online means shortages are likely to persist for quite a while. Recodeal stated that 400G/800G AE... (the excerpt ends here and the remainder was not disclosed). As of now, the companies report no signs of a downturn, and some top-tier customers have already accepted higher contract prices.