It’s interesting how Nokia’s share price drops on the Helsinki stock exchange on quite low volume and starts rising after the NYSE opens, and the same thing happens the next day.
11 Likes
9 Likes
10 Likes
Something approved with Paramount and Warner Bros
12 Likes
Nokia launches Nokia RAN Digital Twin to turbo‑charge AI-native 6G, powered by NVIDIA Aerial Omniverse Digital Twin
Starts in SAN JOSE, Calif. — NVIDIA GTC .. So maybe we’ll hear more then..?
Nokia announced a significant advancement in wireless network simulation with the launch of its Nokia RAN Digital Twin. Built on the NVIDIA Aerial Omniverse Digital Twin (AODT) platform, this system leverages AI and advanced ray tracing to provide physically accurate radio propagation environments for designing and optimizing next-generation networks.
Waiting for other announcements, and specifically the optical side is of interest. Now RAN stuff is more prominent as MWC2026 starts next week.
12 Likes
In response to this Reddit post, here is my brief commentary on the differences between Nokia and Ciena:
Ciena is a pure-play optical player with an order backlog of approximately $5 billion, and it has already been fully repriced by the market as a winner in AI infrastructure.
However, Nokia’s optical business is not significantly weaker. In 2025, its optical margins were comparable, revenue was actually higher, and in the fourth quarter, approximately 30% of the Optical Networks unit’s sales were directed toward AI players and hyperscalers. This is real momentum.
In a nutshell:
- Ciena = Frontrunner → full valuation multiple.
- Nokia = Significantly broader offering (optics + routing chips + in-house chip fab (a new larger fab will be commissioned this year) + in-house packaging + mobile networks + software business + lucrative patents) → complexity discount.
Vertical integration in optical networks is, however, a huge advantage that enables better profitability and the production of more energy-efficient products.
Ciena is currently valued higher than the entire Nokia Group. The valuation gap is striking and, in my opinion, unjustified in light of Nokia’s advantages. Nokia’s investment case is based on growth in hyperscaler accounts and margin expansion, followed by a valuation rerating if the NI unit succeeds in its execution. Ciena is priced for success. Nokia is priced for the need to prove itself.
A more detailed Nokia–Ciena comparison can be found in a separate post.
24 Likes
Well put, that’s exactly how I see it too. Nokia has a golden opportunity in the coming years, and their in-house component-level manufacturing (vertical integration) guarantees better margins than Ciena’s, provided their products are on par with or better than those of other leading companies (like Ciena). As a first taste of this, I believe ICE-X is excellent in its class, combining all the essentials (DSP, PIC, InP…).
14 Likes
10 Likes
Great news keeps rolling in, and we’re only down 3% today. I’m mentioning this just in case there are new investors here and this isn’t familiar territory. 
That Spanish data center deal is a really great thing. It should serve as a reference in the future.
15 Likes
As Nokia’s RAN Digital Twin and NVIDIA’s Omniverse platform merge, I believe it highlights one aspect of AI-RAN development that often goes overlooked in the “AI-RAN vs ASIC-RAN” debate. Even if the industry shifts towards GPU-based AI-RAN architecture, this technology is vital for the development of traditional ASIC-based (ReefShark) hardware. Below is a summary generated by Gemini:
How is Nokia leveraging data and the AI platform?
Nokia has decades’ worth of “Real World RAN” data from hundreds of networks globally. This is utilized as follows:
- Fine-tuning ASIC parameters (ReefShark optimization):
- ASIC chips (like Nokia’s ReefShark) are fixed, but their software parameters and algorithms are adjustable.
- Nokia feeds its collected data into NVIDIA’s AI platform, which runs millions of simulations in a digital twin. This reveals which ASIC chip settings (e.g., power management, memory usage, or timing) perform best in specific load situations.
- The result is features like “Extreme Deep Sleep,” which save up to 25% energy simply by optimizing the chip’s operation based on real-time traffic data.
- Radio Beamforming capabilities:
- Beamforming in Massive MIMO radios involves extremely complex physics.
- Nokia’s digital twin uses NVIDIA’s Ray Tracing to model how radio waves reflect off real-world materials (glass, concrete, trees) in urban environments.
- Through this simulation, AI trains radio control algorithms to direct beams more accurately, which can improve network capacity and spectral efficiency by up to 40%.
- Closed-loop from simulation to production:
- Nokia can test new algorithms virtually in the digital twin before deploying them to physical base stations.
- This significantly shortens the product development cycle (“concept-to-live”) and ensures that ASIC-based hardware gains the benefits of AI without needing to change the hardware itself.
Summary: AI makes “dumb” hardware smart
While NVIDIA pushes GPU-based computing, Nokia’s strategic advantage is the ability to transfer these AI-discovered insights directly into their AirScale and ReefShark portfolios.
This means Nokia is not abandoning ASIC development, but rather making it “AI-native.” The digital twin acts as a laboratory where Nokia’s massive datasets are converted into optimal performance values for ASIC chips and radios.
13 Likes
Probably means that ownership is shifting to the US. In Finland, many are selling now that they finally get their money back?
As a “simple” investor, I had to ask the AI some more questions. Below is a portion, but an important part for my own line of thinking.
An investor’s honest interpretation
If Nokia succeeds in:
Clearly improving energy efficiency
Raising the RAN margin by 2–3 percentage points
Stabilizing cash flow
→ then this strategy creates value.
But it doesn’t make Nokia:
A hyper-growth company
An AI infrastructure winner in the same way as Nvidia
Concise answer
The AI + ASIC strategy is:
Technologically sound
Important for competitiveness
Beneficial for margins
But:
Not a revolutionary growth engine on its own
Does not make Nokia a primary winner of the AI boom
4 Likes
I don’t think that conclusion can be drawn; it’s more likely caused by short sellers and other players. On the lower-volume Helsinki stock exchange, pushing the price down and then buying back cheaper allows for profitable trading.
However, the low trading volume after the price suppression indicates that sellers aren’t exactly available in droves.
7 Likes
Well summarized. We are on the right track and MI will be able to create value for shareholders again in the future. One can only hope that current SoCs offer enough possibilities for dynamic parameterization so that the Digital Twin concept provides a competitive advantage quickly. A new SoC generation takes at least 3-4 years from the drawing board to operational use in the field (SoC + product using the SoC + software adaptation + functional parity).
5 Likes
Nvidia’s Head of Networking Business Gilad Shainer (Mellanox background, one of the keynote speakers at GTC on the topic: Scaling Out and Across: Networking Innovations for Giga-Scale AI Systems) discusses data center networking challenges and Nvidia’s vision as follows:
- Transition from individual servers to the “Data Center is the Computer” model
Shainer emphasizes that in the traditional world, jobs were run on a single CPU or server. In the AI era, workloads are distributed: they require thousands or even hundreds of thousands of GPU units to function as a single entity. The challenge is to connect all of these so tightly that the entire data center behaves like one massive supercomputer.
- Synchronization and Eliminating “Jitter”
The network’s biggest challenge is not just data transfer speed, but synchronization. If, in a 100,000 GPU system, a single GPU is delayed (jitter), all others must wait for it, which ruins efficiency. Nvidia solves this with technologies like InfiniBand and Spectrum-X Ethernet, which are specifically designed to eliminate latency variations and keep computing perfectly synchronized.
- Energy Efficiency and “Co-packaged Optics”
Energy is the biggest obstacle to data center growth. Shainer explains that Nvidia aims to minimize energy waste by moving to Co-packaged Optics (CPO) technology.
• Traditionally, optical transceivers are outside the switch, which consumes a lot of electricity when transferring data to the switch chip.
• In CPO, the optical engine is brought directly inside the switch chip’s package, which can reduce energy consumption by up to 3.5 times and improve reliability.
- “Extreme Co-design” – Seamless cooperation between hardware and software
Nvidia does not see the network, GPU, or software as separate components. Shainer compares a supercomputer to a racing car: every part must be optimized to work perfectly together. This means that even network cables (copper vs. optics) and switches are designed as part of the same “AI factory.”
- Expanding the network to the Edge
Shainer sees the role of the network expanding beyond data centers. The BlueField-4 DPU (Data Processing Unit) plays a key role here: it handles storage, security, and computing optimization both in data centers and at the network edge, enabling AI to be brought closer to the end user.
In summary: From Nvidia’s perspective, the network is no longer just a “pipe”; it is the operating system for the entire AI infrastructure, which must be extremely fast, low-power, and perfectly synchronized.
Gilad Shainer, NVIDIA | theCUBE + NYSE Wired: AI Factories - Data Centers of the Future In this interview, Shainer explains why networking is becoming the “operating system” for AI infrastructure and how it is changing traditional data center architecture.
Gemini: “Gilad Shainer’s views place Nokia’s data center strategy at an interesting crossroads. The ‘Data center is the computer’ thinking emphasized by Shainer and Nvidia’s goal of controlling the entire AI infrastructure create both massive commercial opportunities and direct competitive challenges for Nokia.”
10 Likes
In my opinion, Nokia has “no part or parcel” in the AI boom on the Mobile side, unless there’s a potential competitive advantage from computing capacity being at the edge/low-latency edge computing (e.g., future industrial applications) and the Nvidia partnership/integration, where two respected names join forces when selling packages, for example, to industry. At this stage, I view Nokia’s radios and the Nvidia collaboration mainly as an end-to-end AI enabler from Nvidia’s perspective, and from Nokia’s perspective, at most as potential additional sales to those customers interested in the concept. At this point, I’m more interested in the multipurpose use of radios and the fact that future applications are already being developed. Nokia could suddenly be Nvidia’s end-to-end NW partner at some point, but there’s no doubt about who is calling the shots.
On the mobile side, it remains to be seen how the differences between Ericsson and Nokia will affect margins and whether Nokia will also bring its own purpose-built hardware to the market. Nokia could therefore sell its own solution with a better long-term margin, if potential software updates before the rip-and-replace cycle are taken into account. It could also be that Samsung captures the pot in edge-AI and Ericsson takes the rest! 
We’re living in very interesting times, at least for the next 4-5 years.
On the NW (Network) side, Nokia’s entry means, if successful, mainly that some of the valuations and, of course, the earnings of current AI beneficiaries will flow into Nokia’s coffers as credible competition increases, but overall, a higher number of potential suppliers will lower average margins. On the mobile side, there is suspense over whether this is the new IoT—that we’ll be waiting for 7G before these AI contraptions can be properly monetized.
I also understand Ericsson’s caution. Perhaps it’s not a bad idea to allocate R&D resources only once there appears to be evidence of demand. It’s always cheaper for a follower to develop, and these major disruptions often happen relatively slowly compared to the duration of R&D, so perhaps they’ll only miss the first waves of emerging trends before business really takes off, if those waves even amount to anything.
6 Likes
This has become a great thread to follow, thanks to the active contributors. Reading the latest posts, I reflect on all the energy, and sometimes emotion, that I put into the discussion and maintaining it over half a decade ago now. I was wrong about a lot, but today my portfolio tells me I was right more often than not.
Closely following a single company gives you an edge. I’m no investment guru, but looking at my portfolio over the last five years, things have gone quite well.
What was my point anyway, for taking up your time… Oh right, Nokia is in a situation where massive investments in technology are paying off. Nokia has found its partners and its place at the forefront of technological development. At the same time, it is—once again—metamorphosing into something new.
Nokia is once again interesting, promising, full of surprises, and yet relatively stable. It suits me, and apparently a few others as well.
As for the spring. I don’t believe the surprises will end here.
47 Likes
It undeniably looks like things are finally moving in the right direction for Nokia. We must hope that this is indeed the case, but also that management never sits back and becomes “comfortable” with the situation. Certain companies that are overwhelmingly the best at something can sometimes have a tendency toward exactly this complacency.
Nokia also has a history of disparaging competitors. Rajeev Suri was a master at portraying Ericsson as a loser compared to Nokia in virtually all areas, when ironically it was rather the opposite. I haven’t heard the current management say anything negative about competitors, but rather focus on improving themselves. I sincerely hope it continues that way.
18 Likes
This thread was automatically closed after reaching the maximum limit of 10,000 replies. Continue the discussion in the thread Nokia as an investment (Part 4).