Let’s post part 2 of the dialogue here as well, as I discuss the significance of AI for Nokia in it.
rkurppa said:
A perfectly good answer, but the problem is perhaps that the market values the Nokia Group, not just the Nokia AI & Cloud outlook, which is still a fairly small part of the Nokia Group’s net sales.
Let’s first compare the actual AI and cloud sales in 2025 and 2026:
- Q1: €180m → €350m (+49% with the current corporate structure and excluding the impact of currency exchange rate fluctuations)
- Q2: €220m → €446m (+105% excluding the impact of currency exchange rate fluctuations)
Last year, AI and cloud orders were about €2.4 billion, while over the last four quarters they totaled about €5.4 billion, roughly as follows:
- Q3 2025: €650m
- Q4 2025: €950m
- Q1 2026: €1.0bn
- Q2 2026: €2.8bn
Let’s also keep in mind that the book-to-bill ratio in Q1 was just under three, while in Q2 it was as high as over six. Therefore, significantly more orders have been accumulated than have yet been recognized as sales.
The Q2 order intake of €2.8 billion was likely exceptionally high and should not be seen as a normal level. Instead, a quarterly order pace of about one billion euros seems like a fairly justified assumption: Nokia reached this in Q4 and Q1 and surpassed it sovereignly in Q2. However, the exceptionally strong order intake in Q2 actually suggests that the order trend might be turning towards an even higher level, especially since the order pace for IP networks accelerated in Q2.
This, of course, does not mean that one billion in quarterly orders will translate directly into quarterly sales. That requires time and for the order pace to remain high. Hotard has previously stated that the time from order to delivery in optical networks is typically about 12–18 months and slightly less in IP networks. On the other hand, Nokia stated that about half of the €2.8 billion in orders in Q2 would be delivered over the next 12 months. The current order intake thus provides concrete visibility into future net sales.
If we consider that a quarterly order pace of about one billion is a realistic baseline for the near future and that orders begin to show in sales with a delay, I think it is entirely possible that the quarterly pace of AI and cloud sales will approach the one billion euro level during next year. In that case, we would be talking about an annual net sales of about 4 billion. If a four-billion AI and cloud sales figure is reached in the near future, it would be so significant for the current Nokia that its impact on the whole can no longer be considered minor. And I do not necessarily consider that to be any kind of ceiling in the long term, either.
The market, of course, prices the entire Nokia, not just AI & Cloud. That is precisely why I think it is essential to look at how the entire business composition of Nokia is changing. If AI & Cloud grows from its current small business into a multi-billion euro business in a relatively short time, its significance at the group level will change rapidly. At the same time, based on Q2, the order dynamics in IP networks seem to be improving.
AI-RAN is another trend to keep in mind. Its significance for Nokia in the early stages may be even greater in terms of streamlining the cost structure than in increasing net sales: if moving towards a more software-driven business reduces dependency on proprietary custom chips (and the significant R&D investment associated with them) and increases the share of high-margin software sales, the impact on the currently poorly profitable Radio Networks could be substantial. However, AI-RAN is a long-term project and will only be commercially launched in earnest next year, with volume deliveries possible in 2028, according to Hotard.
Summa summarum: what is currently still a small business with a less than 10 percent share of total sales is, in light of the orders and capacity investments decided by management, turning into a business of significant size for Nokia, where the growth potential is great.
