During Intel‘s (NASDAQ:INTC) most recent shareholder meeting, CEO Brian Krzanich talked about how the company is on track to dramatically grow its silicon-related total addressable market (TAM) from around $ 45 billion in 2013 to approximately $ 260 billion by 2021 — a figure that’s up by around $ 40 billion from what the company said it was back at its 2017 investor meeting.
Of course, it’s not hard to see why Intel would want to advertise such a large TAM — the bigger a company’s TAM, the more growth it can potentially generate should it successfully gain share within it.
Although some of the TAM claims that Intel makes look reasonable, I think a large portion of the TAMs that Intel cites isn’t actually addressable by the company. Here’s why.
Breaking down the TAM
Here’s how Intel breaks down its claimed $ 260 billion total addressable market:
- PC and adjacencies. This consists of Intel’s PC processors, cable gateway chips, Optane solid-state drives, artificial intelligence components, and virtual reality parts (this is likely code for high-performance stand-alone graphics processors). Intel pegs the total opportunity here at $ 60 billion.
- Data center. This consists of Intel’s traditional data center server processors, processors sold into the networking chip market, 3D XPoint-based memory modules, silicon phonics products, and its artificial intelligence efforts. The total opportunity here, according to Intel, is $ 70 billion.
- Non-volatile memory. Intel says that its Optane 3D XPoint memory technology as well as its 3D NAND flash memory (both technologies that can be used in storage drives for both PCs and data center applications) represent a $ 55 billion opportunity.
- Mobile. Intel says that the mobile processor market is a roughly $ 40 billion opportunity.
- Internet of Things. Intel’s Internet of Things opportunity, according to the company, includes automated driver assistance systems (ADAS), industrial applications, video, retail, and artificial intelligence. Intel views this as a $ 30 billion opportunity.
- FPGA. In 2015, Intel acquired programmable logic specialist Altera for $ 16.7 billion. It’s now called Intel’s programmable solutions group, or PSG. Intel says that the total addressable market for PSG’s products is $ 7 billion.
Many of these opportunities are real, but the problem I have is that while Intel is quoting large TAM figures in some areas, its serviceable addressable markets (SAMs) — that is, the portions of the TAMs that Intel can actually go after — are dramatically smaller than the TAMs.
The mobile problem
The most egregious cases of large mismatches between Intel’s quoted TAMs and the SAMs within those TAMs are the mobile and memory figures. Let’s go over those two.
Intel says that the market for mobile processors — that includes both stand-alone cellular modems as well as integrated applications processors for tablet and smartphone use — will be worth around $ 40 billion by 2021. That figure, in itself, is reasonable, but the problem is that Intel can’t actually serve a large portion of that TAM because it doesn’t yet build mobile applications processors that companies can buy. It only builds stand-alone modems and has just one major smartphone customer for those modems — the only smartphone maker that still uses stand-alone modems.
Now, some evidence has emerged that Intel is trying to enter the broader smartphone processor market after it failed to do so in the 2010-2014 time frame, but I think that until Intel formally discloses its ambitions and strategy around that market, putting out a large TAM figure for investors to salivate over (without also including a SAM figure) doesn’t seem like the right way to go.
How big is the memory opportunity?
Intel has been talking a lot about its opportunity in the non-volatile memory market. The company hopes to capitalize on both the secular transition from hard disk drives to faster, more reliable solid-state drives in areas like personal computers, as well as major data centers.
As mentioned above, Intel sees a $ 55 billion memory opportunity ahead that it plans to go after with its 3D NAND flash-based storage drives as well as its 3D XPoint — a type of non-volatile memory that’s much faster than NAND flash but is also more expensive — products.
While, again, I don’t think the $ 55 billion figure is inaccurate, I do think that number incorporates sales of 3D NAND flash technology into segments of the market that Intel doesn’t participate in, like smartphone/tablet storage. So Intel’s actual opportunity is more restricted than the TAM figure would suggest.
A proposal to Intel
I do think that Intel has broadened its total addressable market significantly over the years and the company should advertise that to its current and potential stockholders. After all, more opportunity is clearly better, right?
What I think would be more useful to investors, though, is for Intel to publish SAM numbers. Since the SAM is the portion of the TAM that a company can actually address with its current and future product portfolios, such figures can give investors a much more realistic view of the opportunities ahead of the company.
I think Intel’s SAM numbers would still be quite large (albeit smaller than the TAM figures) and would probably still be enough to keep investors interested in the Intel story.