| | | This earnings call is the aggregation on significant missteps and blunders by the senior leadership of this company.
November 16th 2021 was Cristiano's first major Investor Day as CEO, where he introduced the strategy to move beyond mobile and become the "company for the connected intelligent edge."
It's now 2026 and we haven't seen significant increase in IOT or handset.
| Fiscal Year | IoT Revenue | Year-over-Year Change | | 2021 | $5.06 Billion | +67% | | 2022 | $6.95 Billion | +37% | | 2023 | $5.94 Billion | -15% | | 2024 | $5.42 Billion | -9%* | | 2025 | $6.62 Billion | +22% |
When you compare that to Mobile which has had a quick increase then a more flat stabilization.
| Fiscal Year | Handset Revenue | YoY Change | | 2021 | $16.83 Billion | +61% | | 2022 | $25.03 Billion | +49% | | 2023 | $22.57 Billion | -10% | | 2024 | $24.86 Billion | +10% | | 2025 | $27.79 Billion | +12% |
1) Diversification has been far too slow and management missed significant opportunities. Two such examples immediately come to mind.
First and foremost, AI100 was shipped in 2020 the AI Boom really kicked off in November 2022. The leadership should have had AI200 ready to go in 2024 but instead it was thought up last year and won't be ready till sometime this year.
Nuvia acquisition. When Nuvia was acquired, Qualcomm was finishing up their server SoC. Due to the Arm lawsuit they decided to stop working on the server SoC that was pulled in from Nuvia. Instead of starting over for the server SoC, they decided to start over but for laptops. While this does fit into the diversification strategy it continues in this systematic issue that Qualcomm always defaults to "what it knows" and "what it knows" is not datacenter, but areas that are cyclical and also have the problem of costs.
After laptops instead of going back to a server SoC design they forced the Nuvia team to work on a mobile phone core, then an auto core (which still hasn't shipped) then an IOT core. Only last year did they finally get back to server, and that won't be ready until later. From the documents it may be because Arm was not willing to license TLA cores anymore to Qualcomm under reasonable terms, so if they wanted to ship mobile phones it would have cut into revenues.
This is where I think they still made the wrong call. I would think the revenue of selling Server SoC into this exploding AI market would have significantly outweighed whatever margin cut on mobile would have been licensing expensive TLA cores. It would have also shown the market that the diversification is real, which may have helped with PE expansion.
This isn't the first time that they have dropped the ball in the server CPU. Centriq in 2018 had customers ready to go and due to the hostile takeover Jim Thompson killed the project -- and they never revived it. Another case of a CDMA dinosaur missing the future market.
2) On the automotive side there is real growth going on there, but the numbers just aren't material enough to matter yet. Along with that it's yet another market that relies on HUMANS to buy things instead of Google, Meta, Msft, Oracle, etc with unlimited FCF to spend.
| Fiscal Year | Revenue | YoY Change | | 2021 | $975 Million | +51% | | 2022 | $1.37 Billion | +41% | | 2023 | $1.87 Billion | +24% | | 2024 | $2.91 Billion | +55% | | 2025 | $3.96 Billion | +36% |
On the auto side they've been horribly coy about ADAS, and in particular their language around the perception should point to no real customers outside of BMW. If they existed they would tout them loud and proud.
3) Mediatek is a great example of a company that DID pivot to the correct market by being a cheaper Broadcom.
I'd urge everyone to go look at the investor day 2021 slides it's the same thing that is talked about today the management really is like "oh next year" "oh next year" "oh in 2029". No! they've had 5 years to get this together and they haven't. Why would anyone trust these people? |
|
