Questions
While we have some information, there are still a number of pertinent questions that are unanswered. In many cases, the information would help confirm one or more of the theories for Intel’s approach and the future implications.
Understanding Achronix’s design flow would also be very helpful. What parts of the design were provided by Intel (e.g. reference PLLs), and what parts did Achronix design themselves? What design tools were used, and how did Intel and Achronix cope with any peculiarities from the asynchronous techniques? Also, how does Achronix’s support for QPI relate to the use of Intel’s process?
From the business angle, the structure and terms of the deal would be incredibly interesting. For instance, lower margins for Intel would suggest that they obtain value indirectly (e.g. via complementing products or future acquisition), rather than directly for providing foundry services. Is there an option for Intel to acquire Achronix in the future, or how would the relationship change if Achronix were to be acquired by an Intel customer or competitor?
The most important question though is not about the Achronix deal, but about who else Intel will be working with. It is very clear that they will have other customers, and the nature of those customer engagements would reveal a great deal. A focus around smart phones and consumer electronics would support the idea of Intel seeking complementary technology. A focus on high volume products (as a percentage of Intel’s capacity) would suggest an altogether unlikely interest in the foundry business or perhaps underutilization of the factories.
Conclusions
The reality is that these three different theories are not mutually exclusive, and there may be other explanations as well. There are probably shades of truth to each theory, although to varying degrees.
In the end, any inference that Intel is entering the foundry business should be treated with extreme skepticism. For the technical and business reasons outlined previously, it is an incredibly unlikely scenario – particularly at this point in time with a major foundry market share battle in the offing. Intel has little motivation to explore the foundry business and the barriers are substantial. Intel is firmly wedded to the IDM model, and any departures are likely to be incremental in nature. Moving to a generalized foundry model is too radical of a departure – whereas forming a strategic relationship with certain customers is at least conceivable. There are precedents for this sort of arrangement, Texas Instruments had a long-standing relationship with Sun Microsystems and fabbed their SPARC microprocessors for many years. If Intel is pursuing some foundry business, it is going to be highly selective and exploratory in nature, rather than a fundamental shift; and even that seems rather unlikely.
In contrast, the idea that a manufacturing partnership can accelerate the integration of an acquisition and represents a call option on promising companies seems at least plausible. It is congruent with Intel’s overall IDM business model and is a reasonable way to explore novel and relevant areas with minimal risk. Long term, the FPGA market is interesting, although it’s unclear that Intel would really want to try and compete with Altera and Xilinx. While possible, this seems to be a low probability explanation.
The hypothesis that seems the most likely is that Intel is strategically engaging with companies that complement their current and future product portfolio. The fact that Achronix will have access to QPI seems to imply that it will be paired with Xeon server or embedded processors. There is definitely a niche for FPGA acceleration and Intel already has partners using the older front-side bus. For the future though, QPI is necessary and may be the first step towards on-die integration with a ring (or other topology) interconnect. A tightly coupled FPGA coprocessor is extremely beneficial for high performance embedded applications (e.g. networking and storage) and would also be a good counter to the use of discrete GPUs for HPC. While FPGAs are challenging software targets, they can achieve higher performance in some cases than GPUs and also be used for specialty problems (e.g. deep-packet inspection, cryptography).
The bottom line is that Intel is not likely to enter the foundry business and the most plausible explanation is that they are pursuing complementary technologies. More generally, Intel understands that they cannot do everything for everyone (nor do they desire to do so). This is especially true when it comes to the embedded market, which is nearly impossible to describe because of the variety and breadth of applications. There are already examples of Intel working with third parties to create products that can address different embedded niches. This suggests that Intel may pursue other partnerships in the embedded space to move Atom into new markets, which could actually have a far larger impact than the Achronix deal. In some senses, this would be a second attempt to achieve the same goals of Intel’s unsuccessful partnership with TSMC on Atom – empowering the x86 ecosystem with a variety of hard and soft IP from third parties.
Only time will tell what Intel’s true aims are, as they reveal other customers and perhaps eventually their overall plan, but all eyes are definitely on the embedded space where Intel is keen on expanding their presence.