Why I bought Qualcomm last week
Here is a conversation from the Qualcomm Moderated Thread mesage board between me, Arun and Jeff. Thanks to them for their time and insights.
**********************************************************************************************
Arun’s motivating post:
<The bottom line is what we've often said, Qualcomm's business model is unique, but doesn't belong anymore with the low growth legacy tech companies.>
Historically, by solving the spread spectrum problems with CDMA, Qualcomm built the ecosystem, technologies, and IP that has made huge companies (Xiaomi, Oppo, etc) or made companies huge- Apple, Facebook, Samsung, Google, their growth coming from mobile use.
Despite being a high growth company (by units), Qualcomm fearing monopoly accusation and regulatory overkill from nations has never been able to extract the full value of its contributions. And this has been interpreted by Wall Street that Qualcomm does not really have a moat, and is a commodity player like US Robotics, Netgear etc. In reality, with every generation such as 5G, there are fewer and fewer competitors that can match Qualcomm at the cutting edge of wireless communication. Look at how many have given up or never tried competing over the years - Nokia, Ericsson, Broadcom, Intel, Motorola, Texas Instruments. Apple is still trying. Mediatek has the lower end share. Samsung is still trying. Apple is still trying. Huawei has been forced to stay away by the US Government. Google seems to be wanting to get independent from Qualcomm.
Let us look at what else has happened over the years. Intel was a memory chip maker and then decided to become a microprocessor maker. With the success of IBM compatible PCs, Intel took away a huge chunk of the profits of the business at desktop and server level. The computer revolution depended on Intel improving the performance of its chip to cater to ever hungrier end-user and corporate markets and later data center markets. Intel chips commanded the highest gross margins (partly because Intel was the foundry leader too). AMD was the Mediatek equivalent of the x86 industry, with much lower market share. Later, AMD spun off its foundry business and became a fabless designer of x86 compatible CPU and GPU. So in the microprocessor business, there was the venerable Intel. And also rans. The PC and Server market defined the aristocracy of chip makers. Qualcomm was never considered a serious chip designer.
Then came the digital cellular networks. Between, 1995 and 2005, almost everyone everywhere got a cell phone. And between 2005 and 2021, almost everyone has a smart phone if they could afford one. The ever hungrier applications and functionality demanded by smart phone users (and as defined beautifully by Steve Jobs), has forced the cellphone SOC makers to keep innovating to improve bandwidth, CPU processing power, GPU, all in a small form factor, with lower power use, and quick recharging, while staying compatible with 3G,4G. 4G LTE, 5G, and different bands etc. And Qualcomm is the company that has best met those collective challenges so far. That means Qualcomm has now become one of the best designers of SOCs in the industry. Otherwise, why would smart phones sell for more than most laptops?
Qualcomm of course has not done this alone. Google's Android system (and in future the WearOS) has helped. Having access to the sophisticated foundries of TSMC, Samsung, Global Foundries, and now Intel is a big help as well as lowers the risk. And thirdly, committing to ARM cores, allowed Qualcomm to deliver experiences on phones what Intel could not deliver.
Samsung has foundries and chip design knowledge. After it became a leading cell phone player (thanks to the turn of technology with Qualcomm enabled wireless ecosystem). In the process, it has improved as an SOC provider. So Samsung has become more than a DRAM maker compared to previous competitors such as Micron or SK Hynix. Of course, Samsung has also benefitted from Android and ARM Core licensing to compete.
Huawei was positioned as the next Samsung by the Chinese government. The idea was to invest early in 5G and then challenge or negotiate a better Qualcomm's IP licensing. Huawei also took advantage of Android and ARM core licensing. All that got them to be a serious player. Still behind Qualcomm SOCs, but a player nevertheless. Anyway, US-China trade fights have slowed Huawei down for now.
Now the worldwide Desktop market annually is about $350 billion, and the Server market is about $100 billion. The smart phone market is about $500 billion and already more than the PC and Server markets together. Server market is expected to grow at 8% a year and the desktop at half the rate. Smart phone market is expected to grow at about 9% a year. So now smart phone industry leads not just in sophistication, but also in the dollar value. And the lead is only going to increase with other connected devices.
Let us add another market - the GPU market. It is a component of the PC and Server market. Currently about $30 billion/year. However, because of the growth in gaming, cryptocurrency mining, and AI applications, it is expected to grow at 35% a year. nVidia has the about 80% of the market share, with AMD in the second position. Qualcomm Adreno has a chunk of the GPU market on smart phones, AMD and nVidia are not there. Just note that there are more Qualcomm GPU devices out there than nVidia GPU devices. And Qualcomm has to design these to be more power efficient. And Qualcomm had acquired the Adreno product line from AMD, the second best GPU player. So, Qualcomm's team in GPU is no pushover.
Now Apple under Steve Jobs defined the Smart Phone and Tablet products. Cook is milking them as well as he can. He has developed the ecosystem well with iCloud, Apple Pay, and App Store. But on the technology front, Apple's success so far has been its in-house microprocessor development. Both the A1 and M1 processors developed using ARM cores are cutting edge designs. They are proving that with ARM's help, they can turn the table on previous vendors such as Qualcomm and Intel. (On the other hand, the key members of Apple's design team are now at Qualcomm with Nuvia). Then Apple's products will be continue to charge a premium compared to US, Chinese, Taiwanese, and Korean manufacturers. Next they intend to make a better or equal modem as Qualcomm, and improve their margins further. This is all a defensive play to keep their cash cow customers. Apple still needs to look at much bigger markets - an EV car could be one, as there is room for new players before traditional car vendors recover or new players emerge (Rivian/Lucid/BYD/XPeng/Li/Fisker). Apple will have to outsource the car assembly to an experienced low cost high quality vendor - a Foxconn equivalent in that market. If that is all too late, Apple should try to corner the User Interface part of the car computer system and maybe partner with Qualcomm for the ADAS/Arriver part.
Lets see who the big buyers are of devices with highest end processors. Highest end are the Flag-Ship phone makers and the Cloud servers. There are only 5-6 big flag-ship phone vendors. Apple and Samsung wish to make their own SOCs completely. Xiaomi, Huawei, Oppo, Vivo etc. would love to get rid of Qualcomm, but if they fail to make a competitive product, another vendor empowered by Qualcomm will take their place in the next iteration of phone standards. Everyone is emboldened by having access to ARM cores and think they can catch up.
In the server world, the biggest buyers are governments and the big cloud players - Amazon, Facebook, Microsoft, Google, Alibaba, IBM, Oracle. They will all be gunning for profits being made by their prime vendors - Intel/Altera, AMD/Xilinx, nVidia, and Cisco. They have two ways - make the devices/chips themselves and/or find another vendor to increase competition. They are going to do both in parallel. So there is a possibility that another vendor may squeeze in. Hopefully, Qualcomm combined with Nuvia can reach that level. However, this is less likely at the data center server level so early. As AI processing is likely to dominate the Cloud usage going forward, the Cloud players would like to cut out nVidia GPUs. This will take time, as nVidia GPUs and CUDA language seems to be dominating so far. Again the Cloud players with huge capital budgets, are going to use ARM core licensing to become chip designers themselves. And only the big Cloud players have the clout similar to the large phone makers and big fab-less chip designers to demand Foundry capacity from TSMC, Samsung, or Global Foundry.
So you can see how everyone (except Intel and AMD) is depending on ARM core licensing. And high flying nVidia wants to control that access by buying out ARM. Is nVidia scared or ambitious?
Also, Qualcomm is not getting credit for being a sophisticated chip designer (using ARM cores) - on par with Intel, AMD, nVidia and Samsung. Intel has been extracting over 50% gross margin and big dollars for its chips. Somehow the analysts were surprised on the earnings call with just 35% of gross margin of Qualcomm Technology business. There may be room to grow that margin if Qualcomm starts competing in the desktop and server space. Lets see what Nuvia brings.
-Arun
*****************************************************************************************************************
My comments:
Arun,
<<< Let us add another market - the GPU market. It is a component of the PC and Server market. Currently about $30 billion/year. However, because of the growth in gaming, cryptocurrency mining, and AI applications, it is expected to grow at 35% a year. nVidia has the about 80% of the market share, with AMD in the second position. Qualcomm Adreno has a chunk of the GPU market on smart phones, AMD and nVidia are not there. Just note that there are more Qualcomm GPU devices out there than nVidia GPU devices. And Qualcomm has to design these to be more power efficient. And Qualcomm had acquired the Adreno product line from AMD, the second best GPU player. So, Qualcomm's team in GPU is no pushover. >>>
Thanks for a nice overview. Don’t know if this has been posted before but I found it to be quite informative.
Meet the AI Powering Today’s Smartest Smartphones
Message 33556512
Cheers,
Frank
******************************************************************************************************************
Matherandlowell,
Me must respectfully disagree.
<<< Nvidia now trading at 55x estimates for 2024. >>>
Yes, but the main areas of growth for NVIDIA, data center AI chips and robotics AI chips and software, including autonomous vehicles, are growing at a five-year CAGR of nearly 40%. At these rates the markets will grow five-fold in five years and NVIDIA will be a prime beneficiary.
<<< From Barron’s: Nvidia which will report earnings next Wednesday—is primarily a maker of graphics processing units (GPUs), which were originally designed for applications in gaming and film. Increasingly, GPUs are used in high-performance computing applications, such as running data centers and powering artificial intelligence (AI). >>>
From Forbes: NVIDIA Is Not A Chip Company. It’s A Platform Company. Message 33570145
NVIDIA’s state-of-the-art software platforms in autonomous driving (Drive), robotics (Isaac and Omniverse), and Healthcare (Clara Holoscan) all complement their AI chips and their new NeMo Megatron framework for training trillion-parameter language models and their Triton Inference Server offers multi-node distributed inference features for new domains and languages is state of the art. In addition, their legacy gaming GPU’s will continue to contribute nearly 50% to revenues and will grow substantially next year as industry-leading GeForce RTX 30x graphics chips become more available as the chips shortage moderates; they are planning significant price increases increases for their RTX 40x series next year.
My money is on NVIDIA to continue it’s phenomenal growth for the next five to ten years, at least. JMHO!
Cheers,
Frank
******************************************************************************************************************
Jeff’s response:
Qualcomm is not a chip company, it's a platform company. A similar article could have been written about Qualcomm, featuring their software, AI, robotics, auto, IoT, drones, health care, and smart cities/factories ecosystems. They are also a mobile connectivity, GPU, DSP, CPU, and modem-to-antennae RFFE company. Soon, Nuvia improvements on the CPU side, as well as the company's long cultivated research and development efforts, will begin bearing fruit in mobile computing, edge computing, and other growth endeavors.
Clearly, Nvidia and Qualcomm are finding their respective visions converging and intersecting in multiple growth markets, while AMD and MediaTek, respectively, try to take share in their legacy businesses.
Nvidia is a great growth story, and Jensen is the consummate founder/salesman. It appears that Qualcomm has finally determined that its messaging needs to be improved, to more accurately reflect its current and future business realities.
Tuesday will bring a major effort to reshape the way the investment community perceives Qualcomm. After all, as Jensen has reminded us, perceptions are reality.
***********************************************************************************************************
My rejoinder to Jeff:
Thanks Jeff,
Your recommendation and comments as well as Arun’s and the articles I’ve read have convinced me to make a modest investment in Qualcomm on Monday, before the hoopla on Tuesday. I am continuing to diversify my holdings in AI & Robotics, which now will include NVIDIA, Google, and AMD in the growth arena, and Baidu and Qualcomm in the value arena. I also own a bit of the Robotics and AI ETF IRBO, which I’m considering liquidating after year-end since it hasn’t performed as well as the others (the strong companies get diluted by the weak companies). This is a investing theme which is becoming a bit too large a part of my portfolio than I find comfortable but I have become convinced of the long-term growth story. AI, AI chips, Robotics AI, and Robotics software are all forecast to grow at about a 40% CAGR over the next five years, growing five-fold in this time. The companies in my portfolio all appear to me to be the world leaders and will benefit from this extraordinary exponential growth.
Cheers,
Frank
************************************************************************************************************
Jeff’s rejoinder to me:
Hi Frank,
From the list of stocks on your profile, it's obvious that managing your own retirement portfolio was a superlative move. Congratulations! I, too, am an investor in Nvidia and AMD, though I admit that I was late to both parties.
Thanks for sharing your insights with Qualcomm investors, as well. Since the end of "the bubble", QCOM has been languishing as a value stock, fighting one legal, regulatory, and vertical-aspiring customer battle after another, while paying large dividends to us early investors for the patience to wait a seemingly interminable 20 years for what comes next.
The company has been greatly de-risked since settling with Apple, and winning the FTC v Qualcomm case in the 9th Circuit Court of Appeals. They are making the most of this period of peace, and you enter the stock at an interesting time - priced as a value stock, but growing both nominally and relatively, at rates greater than AMD, and similar to Nvidia.
My hope and expectation for us all, is that QCOM will again be perceived as a growth stock.
Best of luck,
Jeff
************************************************************************************************************
My comments:
Thanks Jeff,
Your recommendation and comments as well as Arun’s and the articles I’ve read have convinced me to make a modest investment in Qualcomm on Monday, before the hoopla on Tuesday. I am continuing to diversify my holdings in AI & Robotics, which now will include NVIDIA, Google, and AMD in the growth arena, and Baidu and Qualcomm in the value arena. I also own a bit of the Robotics and AI ETF IRBO, which I’m considering liquidating after year-end since it hasn’t performed as well as the others (the strong companies get diluted by the weak companies). This is a investing theme which is becoming a bit too large a part of my portfolio than I find comfortable but I have become convinced of the long-term growth story. AI, AI chips, Robotics AI, and Robotics software are all forecast to grow at about a 40% CAGR over the next five years, growing five-fold in this time. The companies in my portfolio all appear to me to be the world leaders and will benefit from this extraordinary exponential growth.
Cheers,
Frank
************************************************************************************************************
AI & Robotics Expected Growth
One of the reasons I’m so excited about the AI & Robotics field is the enormous expected exponential growth of this market. AI, AI in Robotics and Robotics software markets are all forecast to grow at nearly 40% five-year CAGR. At this rate these markets will grow five-fold in five years. See market forecasts below.
I am a demizon of and post extensively on the AI, Robotics & Automation message board moderated by my friend Glenn Petersen. He has recently completely updated and revised the Introduction Header to the message board. Definitely worth reading. If you’re interested in the field and players, visit at the following link:
Subject 59856
Cheers,
Frank
************************************************************************************************************
Comment from Bill Wolf:
Qualcomm and BMW Group to Extend Their Long-Lasting Technology Collaboration to Automated Driving
- BMW Group Selects Vision Perception, Vision System-on-Chip and ADAS Central Compute System-on-Chip from Qualcomm Technologies' Snapdragon Ride ADAS Platform for its Next Generation of ADAS and Autonomous Driving Systems -
finance.yahoo.com
************************************************************************************************************
My comments:
Re: I heard $100B TAM potentially growing to $700B over the next decade.
I calculate that this is a 24.5% CAGR for ten years. Pretty impressive.
I am not an engineer but I liked Jim Thompson’s presentation.
I like how Christiano tied in the Metaverse! I thought both Jim’s and Christiania’s presentations were very good at emphasizing that Qualcomm is no longer just a smartphone company.
Surprised Qualcomm’s share price is not reacting positively? I want to double down on my Qualcomm position so this gives me the opportunity.
Cheers,
Frank
************************************************************************************************************
My comments:
Qualcomm Growth, Current and Future
From presentation by Akash Palkhiwala, CFO of Qualcomm.
As far as the future, Akash said that automotive SAM would grow from $3B today to $15B in five years, a CAGR of 36%, and that automotive revenues would grow from $1B now to $3.5B in five years, a CAGR of 28.5%.
************************************************************************************************************
My comments:
Doubled down on QCOM. Now I have 200 shares. Should have bought more Monday but in long run won't matter. Exponential growth trend seen continuing for five years at least. Thanks to Arun and Jeff for convincing me to invest in QCOM.
Cheers,
Frank
|
|
