I cannot stress this enough, but there is a major upside catalyst coming for Intel (NASDAQ:INTC) on July 22. When I checked Intel's investor relations site for any upcoming events, I saw the following:
Click to enlarge images.
While this seems innocent enough, and for most investors it may not appear to be of consequence, I believe that on this webcast Intel will be announcing its next generation micro-server and networking infrastructure parts based on the Silvermont core known as Avoton and Rangeley, respectively. I believe that many investors are completely unaware of the magnitude of these launches, particularly as the sell side has remained mum in light of their promotion of ARM Holdings' (NASDAQ:ARMH) data-center aspirations.
These launches will dramatically change the investing landscape, and I will detail why in this article.
Micro-Servers: Intel Will Be First to Market With Most Advanced Technology
While the Wall Street pundits and sell-side analysts are busy getting excited over the fact that ARM's next-generation chips will be 64 bit, they seem to forget that there's so much more to winning the data-center sockets than simply supporting tons of memory. For example, Imagination Technologies' (OTCPK:IGNMF) MIPS has included a very elegant 64 bit instruction set architecture and a couple of well-designed stock processor cores. Additionally, network players such as Cavium Networks (NASDAQ:CAVM) and Broadcom (BRCM) have been designing custom MIPS64-based cores for ages now. Yet despite the ubiquity of MIPS64 core designs as well as a clean, elegant 64 bit instruction set architecture, none of these emerging micro-server players had previously tried to go up against Intel in the server space. And for good reason.
In the server/HPC space, Intel managed to muscle out almost every processor architecture, with the remaining IBM (NYSE:IBM) POWER and Oracle (NASDAQ:ORCL) SPARC relegated to increasingly niche niches. Intel has largely won this battle as its designs have been both cost effective (leveraging Intel's massive IP and scale in the PC space) as well as leadership in terms of performance/watt.
Now, the major fear on the Intel side is that the ARM vendors would do to Intel what Intel did to IBM/Oracle/everybody else. However, I believe quite the opposite is true, and that the battle will actually be easier for Intel this time around as it merely needs to kill the efforts from small start-ups and firms with 1/100th of Intel's R&D budget at birth. At the same time, Intel has established a rather wide software moat, meaning that even if Intel had a slight performance/watt deficiency, the cost of porting over the software and making sure it plays nicely with the rest of the data-center elements makes mixing and matching architectures within the data center (which will likely have plenty of Intel chips for the higher performance workloads) a costly proposition from a TCO perspective.
However, I do not believe that there will be any need for Intel to fall back on the software ecosystem advantage; I believe that the firm's upcoming Avoton micro-server part will be extremely competitive with any of the competition's parts on a performance/watt basis.
Avoton -- Crashing Everyone's Party
Avoton is Intel's upcoming micro-server part, and the first of two parts that I believe will be announced at this Intel webcast event. This will be based on eight Silvermont cores and feature extensive integration of I/O, networking fabric, and so on. This will be built on Intel's 22 nanometer SoC process, and should -- with a combination of best-in-class design and an inherent transistor leadership -- offer significantly better performance-per-watt than the competitive solutions from either Applied Micro (NASDAQ:AMCC), which will be built on TSMC's (NYSE:TSM) dated 40 nanometer process, or AMD's (NYSE:AMD) own ARM Cotex A57-based solution built on TSMC's 28 nanometer process.
Furthermore, note that Applied Micro's part is scheduled for launch in Q1 2014, which gives is a significant time-to-market disadvantage. Coupled with a highly delayed, unproven design from a team that has yet to demonstrate that it can successfully execute a product of the scale that it talks about, I see very little chance for success against the Avoton parts in winning any major design sockets. Also, Applied Micro is not likely to have the scale in order to effectively compete on price, which is what I expect will be necessary given the technological disadvantages that the firm faces. Also, Applied Micro's constant delays of its X-Gene product is likely to make potential clients wary of Applied's ability to execute to its road map for future products.
AMD, while a much more proven name in the server space, is also likely to face challenges against Avoton. First off, commercial availability is planned for the second half of 2014, which means that it will not be competing with Avoton at all, but instead with whatever 14 nanometer successor that Intel puts out. Trying to compete with the 22nm Avoton part would have been tough enough, but with a 14nm Airmont-based SoC it is tough to see how AMD can have a meaningful competitive advantage in this area. AMD has, at least, a larger revenue base so it can afford to try to compete more effectively on price, but it still seems like it's going to be an uphill battle in light of Intel's technology and scale advantages.
This Launch Is a Double-Edged Sword
While it seems as if this should be well known, I remain convinced that the Street and the sell side alike are unaware of the massive long-term implications of this launch. As threats of market share loss in servers evaporate, I expect Intel to see renewed optimism and perhaps a good bit of multiple expansion.
However, this is kind of tricky. The launch occurs just five days after the earnings report, which means that Wall Street will pretty much have made up its mind on Intel the night before the launch. If Intel does well in Q2 and guides nicely for Q3, then the sell side is likely to spin the micro-server opportunity as "incremental" to the firm's DCG revenues and whatever post-earnings positive momentum should extend. But if Intel's Q2 stinks or if the Q3 guide is uninspiring, then the sell-side analysts will try to spin something about how Intel's margins will go down (despite the fact that even Applied Micro expects 70%-plus margins on its parts) or something about how AMD's SeaMicro fabric may be better, etc. That's how this game is played.
So, I think that this launch essentially means that there's going to be incredible amounts of volatility in any stock involved in micro-servers, and in particular Intel. My long-term view is that these parts help Intel grow its data-center revenues substantially and will be effective at diffusing the ARM "threat," but in the near term, as this launch occurs after the earnings report, it will simply serve to introduce further volatility.
I know where my bets are on both Q2 and this launch, but what do you think? Tell me in the comments how you plan to play this.
Disclosure: I am long INTC, AMD, ORCL. I may go short ARMH or AMCC at any time. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.