I realize that most people opening this article will be shocked to see that I am the author of a title that seems glowingly positive about Advanced Micro Devices (AMD 1.33%). You're probably expecting some "bait and switch," coupled with a detailed look at why you shouldn't fall in love with AMD.

Nope. None of that here. In this article, I'd like to present what is more or less the bull case for AMD from the perspective of somebody who is admittedly fairly bearish on the stock.

The upside, if AMD succeeds, could be tremendous
AMD's market capitalization is about $3 billion today, and if you believe the current sell-side estimate, the company is on track to generate $5.96 billion in revenue, which -- at a glance -- puts the company at a rather inexpensive looking price-to-sales ratio of 0.52. This is seemingly dirt cheap if price-to-sales is your go-to metric.

Things don't look as rosy if you look at the company's price/earnings against current consensus, which sits at $0.19/share. This means that, on a forward basis, the stock trades at about 21.4 times this year's earnings, and about 17.6 times 2015E consensus. At a glance, this doesn't really look dirt cheap, but it's nowhere close to the most expensive semiconductor player on the planet. (Note that these are non-GAAP numbers, but we'll brush that aside for now.)

That said, it's hard to justify, say, a $5, $6, or even $7 stock price based on these current estimates. Being long AMD is a bet that these estimates are too low. So, if you're an AMD bull, you probably see a good deal of upside to current estimates. The question, then, is where could it come from?

Taking a look at the moving parts of AMD's business
The following slide from the company's annual stockholder meeting illustrates fairly well where management sees the growth opportunities:

Source: AMD.

Graphics to the rescue?
In professional graphics, AMD is in competition with rival NVIDIA (NVDA 3.71%), which holds the lion's share of the market (it's a roughly 20/80 split ), so if AMD can gain share there without completely crashing its margins to do so, this could represent a fairly lucrative opportunity. At a recent conference, AMD indicated that the company has been aggressively hiring in order to bolster its capabilities there to gain share.

Game consoles only the beginning for semi-custom, it seems
AMD is also counting on semi-custom deals, such as the contracts it won to do the game console chips for Sony/Microsoft. AMD executed very well with that project and, as a result – according to AMD – this has sparked intense interest from other clients looking for semi-custom products. While it's really difficult to imagine that there are too many individual contracts as large and lucrative as the game console deal for AMD, enough of such deals in the aggregate could move the needle.

Embedded and dense server could potentially drive upside
Next -- and these should probably be lumped together -- is the company's focus on "dense server," or micro-servers as some call them. AMD is attacking this market with designs that will come in both ARM and X86 flavors depending on what the customer wants.

While the size of this market is unclear, and while AMD appears to be one of the leading micro-server players, the idea here is that the company could -- by playing here -- improve its server share dramatically from current sub-5% levels. It also looks as though the embedded designs and dense server chips will be very similar, per AMD's recent roadmap.

Ultra low-power client
Finally, AMD's position in the client market during the last few years hasn't been particularly good, but AMD is promising both an Android strategy through the utilization of ARM-designed cores, as well as a brand-new X86 core for the traditional Windows PC market. While it's unclear how much share AMD can take, the idea, once again, is that the TAM expands, which, with proper execution, should lead to revenue upside.

See why it's easy to fall in love with AMD?
There's no denying that if AMD can capture meaningful professional graphics share, continue building up a robust pipeline of semi-custom deals, grab real share in the server market, and revitalize its story on the client side of things, it should see some pretty dramatic revenue upside. Indeed, if we do some back-of-the-envelope math, working under the following (admittedly optimistic) assumptions:

  • AMD can grow total sales from ~$6 billion/year -> $8 billion/year
  • AMD keeps gross margins at roughly 35% (this is conservative since this assumes no leverage)
  • AMD keeps operating expenses at roughly $1.7 billion to $1.8 billion per year

This would ultimately lead to $1 billion/year in operating income. Applying an Intel-like 10 times operating income multiple, and netting out the approximately $1 billion net debt position that the company has (which would get much easier to pay off at this run rate), we get a business that could command a market capitalization of approximately $9 billion or, at the current share count, about $12/share.

Now, growing sales by 50% isn't going to be easy, and keeping operating expenses flat in support of such growth is probably not a realistic assumption; but it is clear that the bulls are expecting a rather large payday if AMD can actually execute. But that's the million dollar question, isn't it?