I remember when Intel (INTC) was hyping up the fact that it had "won" the Samsung (OTC:SSNLF) Galaxy Tab 3. At first, as an Intel bull, I was encouraged and believed that this had signaled that some very good things were in the pipeline for Intel over at Samsung. While the following development probably doesn't invalidate the long-term thesis that Intel will make some inroads over at Samsung, it does illustrate a failure to execute on Intel's part.
The New Galaxy Tab Pro 10.1 - Powered By Either Snapdragon Or Exynos
According to VR-Zone (which was astute enough to notice this tablet pop up in the AnTuTu benchmark database), the next generation Samsung Galaxy Tab Pro 10.1" (which seems to be a replacement for the Galaxy Tab 3 10.1") packs an Exynos 5420 (Samsung's own home-grown silicon). It is likely that, in as similar fashion to Samsung's higher end Galaxy Note products, that any edition with cellular connectivity will sport a Qualcomm (QCOM) Snapdragon.
It doesn't take analytical genuis to understand why Intel didn't have a snowball's chance in hell of winning this new socket: it didn't have a part ready (no Android for Bay Trail yet). Further, the only part that could have even been remotely compelling for this market - the company's Atom Z3770 ("Bay Trail") - is competitive with the Exynos or Snapdragon on CPU performance, but still lags the Mali-T628 in Exynos 5420 or the Adreno 330 found in Snapdragon 800. In short, Intel still doesn't have a solution good enough to win high end sockets on performance per watt merit.
As An Intel Investor, I'm Irked, But Not Surprised
This isn't a surprise to me as somebody who is knee-deep in the Intel story. I know that Bay Trail still isn't good enough to win the really important "hero device" sockets, and frankly, its really disappointing that Intel executed so poorly on getting Android for Bay Trail ready to go. Intel could have built upon the momentum that it had gotten rolling with the Galaxy Tab 3, but instead they blew it.
That being said, Bay Trail is not too far off the mark from being a "leadership" solution and with the appropriate subsidies/contra-revenue, Intel will be able to take some real market share with this part. It's not a bad chip, and if Intel's Cherry Trail is as competitive as the leaks suggest (and if Broxton is as competitive as little birdies have hinted to me), then Bay Trail will be a good enough "placeholder" until the big-guns arrive.
Let me be clear. Bay Trail is leagues better than the Clover Trail/Clover Trail+ parts that Intel tried to sell (they were perhaps a year or two too "late"), and it is good enough provided that Cherry Trail/Broxton look absolutely incredible to the OEMs (not an outlandish proposition). Nobody will design Intel in if they believe that Intel will keep missing the mark time and again, but I think with Broxton, Intel finally takes decisive performance per watt leadership at a cost structure that allows this venture to be profitable. Cherry Trail looks compelling, too (4x "Airmont cores @ 2.7GHz, 16 EU Gen 8 GPU), but Intel didn't tout it as an "absolute leadership" product as it did Broxton at its investor meeting, leading me to be a bit more cautious.
Intel can do it, but it really needs to get its low power SoC teams to start executing better. Bay Trail should have been out in early 2013 if it wanted to be a true leadership product, and - quite honestly - Intel could have done much more with its 22 nanometer process. At any rate, it's frustrating that Intel didn't have the goods to win a next generation, high-end Galaxy Tab, but I think investors need to understand that Intel isn't quite in a position to win these ultra high end hero devices just yet.
Things get a whole lot better in late 2014 with Cherry Trail and mid 2015 with Broxton, but until then, the spotlight will be on just what sockets Intel can brute-force with Bay Trail so that we know what kind of designs Intel's next generation products will go into profitably. It's a long-term game, but I think with patience, Intel will finally begin to deliver and, in doing so, finally yield returns in excess of the broader market.