Qualcomm Investor Day – Pick and Mix

4 hyperscalers in the bag.

• With 4 hyperscaler contracts and a large upgrade to long-term guidance, the issue of a weak handset market and losing Apple as a customer is rapidly becoming an irrelevence.
• Qualcomm is also doubling down on its long-term approach of offering the entire technology stack but allowing customers to pick and choose the bits that they want, which I think is going to work especially well in the data centre.
• Qualcomm brought everyone to New York and unleashed a torrent of information, where the most important messages were:
  • First, 2029 financial targets: where non-smartphone revenue target has been upgraded by 82% to $40bn from the $22bn set in 2024.
  • This is primarily driven by the new data centre business and upgrades to automotive, and sets the company up for growth to accelerate materially in fiscal year 2027, which starts in October.
  • An adjusted EPS target of $18+ per share was also given, which is 25% above the current consensus estimate of $14.45, which looks low to me.
  • This is because the company thinks that operating leverage will allow it to take 300bp – 400bp out of OPEX as a percentage of sales, meaning that margins should increase.
  • A 2029 forecast of $18 per share implies some margin degradation, and so I think that this number could easily be beaten.
  • If one puts a 20x PER multiple on the 2029 EPS target, one gets to $360 per share, which looks pretty reasonable to me on a 3-year time horizon.
  • Second, Data centre: which was the star of the show with 4 product lines, a new memory architecture and 4 hyperscalers on board as customers.
  • Meta and Microsoft were willing to announce that they are customers, while two others declined to be identified.
  • I suspect that one of these is AWS, while the other one could be xAI or Oracle.
  • Revenues are expected to be $5bn in FY2027 and $15bn in FY2029 as the four product lines come into production over the next 2 to 3 years.
  • The view has long been that, as long as Qualcomm could secure two hyperscalers on top of Humain (first customer), then the business would be viable.
  • With 4 hyperscalers, Qualcomm is now a credible player in the data centre setting it up for sustained growth over the next 5 to 10 years.  
  • Third, Modular and HBC: are two areas of particular differentiation in the data centre.
  • Modular is a software platform for the data centre that Qualcomm is acquiring for around $4bn in shares.
  • It is an alternative to Nvidia’s CUDA, but supports every hardware vendor in the market, including Nvidia, and Qualcomm claims that the performance using Modular is on par with CUDA for inference.
  • If this proves to be accurate, this will help open up the data centre market to other vendors.
  • Modular will remain independent, reporting directly to the CEO, not the data centre business, which will be crucial to get others to use it.
  • HBC is a new memory architecture where Qualcomm has worked out how to put the accelerator directly underneath the memory stack without producing excessive heat.
  • Memory remains the main bottleneck in the data centre, and if this works as promised, then it will go a long way to alleviating the bottleneck, allowing a large improvement in token production per watt and per $.
  • Fourth, Smartphones: which continue to have a tough time but are becoming less and less relevant.
  • The memory shortage is causing prices to rise, which means that shipments will weaken, which will cause some problems in the short-term.
  • The target here is to keep revenues flat through higher content per phone as AI agents start to run on edge devices, which should help offset any weakness in volume shipments.
  • This business will be less than 50% of revenues in FY2027 and a third of revenues in 2029, in a clear demonstration that the diversification strategy is working.
  • Hence, I think that the days of smartphone dynamics driving the share price of the company are coming to a close.
  • Fourth, Automotive, industrial and robotics: which were the icing on the cake.
  • The recent acceleration of growth seen in FQ2 2026 in Automotive looks set to continue, and the FY2029 target has been upgraded from $8bn to $10bn.
  • Stellantis is now on board for both infotainment and ADAS, and Qualcomm will soon become the largest supplier of automotive semiconductors in the market.
  • PCs, XR, Industrial and robotics all remain a work in progress or development, but the signs are good.
• The net result is that Qualcomm did precisely what it needed to do at its analyst day, and everyone will now be having another look at their spreadsheets, meaning that FY27, FY28 and FY29 estimates are now likely to rise.
• Most importantly, Qualcomm is sticking to what it is good at, which is executing on time and offering a horizontal solution.
• This means that it offers everything one needs to make a digital device, but the customer can pick and choose which bits it wants to buy and which bits it wants to make itself or buy from someone else.
• Given how the hyperscalers are all working on their own silicon and software, I think that this approach, which has worked extremely well in smartphones and automotive, is just as applicable in the data centre.
• Hence, I am pretty confident that the 2029 forecasts will be met or exceeded, which in turn means that the shares have much further to run.
• Qualcomm is pretty much the only semiconductor supplier that has not added multiples to its market capitalisation as a result of AI, and perhaps its turn has finally come.
• I don’t think it is going to bounce 4x in a month, as Arm did, but I am looking for a sustained and steady return over the next few years.
• I have a position in Qualcomm and remain very happy to sit on it.