When Intel (INTC +1.07%) launched its first 22-nanometer tablet processor, which brought a brand-new design and promises of leadership, it seemed the difficulties Intel had been facing as a result of tablet cannibalization of the PC may have been coming to an end. A strong presence in the tablet market, coupled with world-class factories that gave Intel a cost-structure that seemingly nobody else could match, Intel looked poised to do very well in the years ahead. Unfortunately, reality wasn't so kind to Intel.
Poor graphics performance, lack of integration, and bill-of-materials problems plagued Bay Trail
The Bay Trail-T part from Intel suffered from three principal shortcomings with respect to chips from the likes of Qualcomm (NASDAQ: QCOM), MediaTek, and Allwinner:
- Lack of Integration -- While competitors began integrating cellular and connectivity capabilities into some of their chips (particularly for the mid-range and the low end), Intel's chips required external chips for quite a number of functions. This is more common at the very high end, but that's the next problem.
- Inadequate performance -- For all of Intel's song and dance about the superiority of its architecture and its manufacturing capabilities, Bay Trail-T offered meaningfully worse graphics performance than its rivals such as the Qualcomm Snapdragon 800 and anything featuring PowerVR Series 6 graphics. The CPU performance was good, but not Earth-shatteringly so; and
- Bill of materials issue -- At the platform level, not even taking into account the lack of integration into the chip itself, Intel ran into a significant problem with the bill of materials required to support the Bay Trail platform. According to Intel CFO Stacy Smith, this bill of materials disadvantage relative to competitors is on the order of $20 -- clearly unacceptable.
The bad news is that as a high-end/leadership part, Bay Trail just wasn't going to cut it. Even if we assume it had leadership CPU performance, the graphics performance of this part was simply lackluster compared to solutions from Qualcomm and others that were available on the market even before Bay Trail's launch. What was Intel to do?
Flood the low end; be a big fish in a small pond
At the November Investor Meeting, Intel's management team indicated that they wanted to "go broadly" into the mainstream/value portion of the tablet market with Bay Trail in order to build a footprint for its architecture and to get developers on-board. Intel also said that in order to drive those volumes with Bay Trail, it would offer contra-revenue to offset the expensive bill of materials that the platform requires to make the platform more attractive.
While getting developer/OEM support is key, and while this will be a very good strategic use for the chips, it is likely this decision was made after it became clear that Bay Trail was simply not very competitive at the high end in all of the vectors that mattered. A good save on Intel's part, but it's unfortunate that Intel is still not quite hitting leadership performance, even with a process lead and seemingly unlimited R&D resources.
Foolish bottom line
Intel is the world's leading semiconductor company by revenue, and its execution in the PC market as well as the server market has been nothing short of phenomenal, but mobile is still very much a work in progress. The hope is that by 2015, Intel has leadership products across the board, but this is what Intel investors expected in 2013, and then again in 2014. At this point, Intel needs to deliver something unequivocally competitive -- no excuses, no caveats -- before the market will truly begin to view Intel's mobile efforts as viable.
