Nokia's real draw is indium phosphide and data center interconnect, not AI-RAN, which is pure hype. It already won Google's DCI deal because Ciena is out of capacity, and it should secure more deals from other cloud providers, possibly Amazon. x.com
3X Long Labubu
Long-Term Investor, Hobby Trader
Nokia's agentic-AI work with Nvidia adds another bullish development to its network-automation case. Nokia says their models and network intelligence help operators act on insights while keeping control of sensitive data. L&T Technology Services also announced engineering and R&D work with Nokia in India; after strong Q3 results, the shares still look ridiculously priced. x.com x.com
Nokia is well positioned for data centers' shift from copper to optics because it makes and packages indium phosphide optical chips in-house, and could eventually double. Through Infinera it owns scarce InP fab capacity, is expanding San Jose 20-fold and buying NXP's Chandler campus, and grew US photonic packaging tenfold, avoiding Coherent or Lumentum markups. Q2 AI and cloud sales rose 105% on EUR 2.8B of orders, with backlog at 6.3x quarterly segment revenue. Curbs on Chinese optics help. … Show more
Nokia looks overhyped for now, and a weak Q3 could bring a short-term pullback despite a bright long-term setup. Management said Q3 will be flat because high-margin software revenue was pulled into Q2, and restructuring charges of €390M still weigh. AI and cloud revenue rose 105%, with €2.8B of binding Q2 orders, but memory and optical supply limits are capping growth. Q4 needs to confirm the thesis before buying, likely toward year-end. x.com x.com x.com
Nokia's AI and cloud orders are set to turn into revenue, with about half of Q2's EUR 2.8B intake expected to convert over the next 12 months. Its San Jose fab ramps in Q4, Pennsylvania test and packaging capacity expands 10x from Q3, and an NXP Chandler fab lease for indium phosphide chips starts in early 2027, all for demand expected in 2027-28. Nokia targets EUR 2.7-3.2B of comparable operating profit by 2028, up from EUR 2.1-2.6B guided for 2026. x.com
Nokia is turning into an AI network powerhouse, and its fundamentals and technicals are aligning for a major move. Its $1 billion Nvidia partnership on AI-RAN and 6G is entering live commercial trials with operators this month, a Microsoft expansion links Nokia Data Suite with Microsoft Fabric, and it is converting a €2.8 billion AI order book while winning AI-native 5G deals with Taiwan Mobile and Chunghwa Telecom. Price has reclaimed its 50 and 200 EMAs; a move above 12 could reach 13+. x.com
Nokia does not deserve the roughly 50x multiples of optical pure plays like Ciena and Coherent, because it still carries large, shrinking legacy businesses in wireless, PON and routing. This rebuts Citron Research's case that comparisons with those peers make Nokia a $20 stock. x.com
Nokia stands to benefit from optical networking demand: Ciena's demand commentary at its investor forum reads positively for Nokia. Rosenblatt sees clear upside for Ciena following the event. x.com
Nokia is becoming a credible Western AI infrastructure pick, though only its Network Infrastructure half. The Infinera deal added in-house photonic chip fabs and hyperscaler access, Nvidia's $1B stake brings AI-RAN co-design, and the segment grew 23% in FY25 with EUR2.4B of AI and cloud orders. Licensing at near-100% gross margin and EUR4.9B of net cash add support, while Mobile Networks is a commoditizing anchor. Risks: AI capex peaking before Nokia wins optical share, and synergy slippage. … Show more
Nokia's guidance raise was just an accounting shift, leaving a steep climb to its profit target. The bump to €2.1-2.6 billion came from moving two business lines into discontinued operations; JPMorgan asked why operating targets weren't raised. With about €735 million of H1 operating profit, Nokia needs €1.8-1.9 billion in H2 for the top end, so delays risk a miss. Memory and optics remain bottlenecks, and only about half of €2.8 billion in AI and Cloud orders converts within 12 months. x.com … Show more
AI-on-RAN and near-AI compute, feeding robots and inference that need low latency in specific locations, look like a better bet than specialized silicon, which favors Nokia and AmpliTech. x.com