Intel Corp (INTC)

[wbmw]

Intel also sees good progress in reducing mobile losses by $800m this year. Key drivers being smaller BOM differential in Sofia LTE and lowered investments in mobile due to partnerships with Rockchip and Spreadtrum.

>> That's still represents a loss of $3.6B, as Intel is losing $1.1B/q currently.

I think the quote was originally supposed to imply that by Q4 (the end of the year), the quarterly loss target is to have shrunk by $800M. In other words, they expect to have a quarterly loss of $300M, instead of $1.1B, by the end of the year.

That's the way I read it, anyway.

Any gains SoFIA makes will be muted for Intel by the fact that they only take a slice of the already cheap, low-end processor. TSMC and Rockchip will take their cuts too.

TSMC will lose their cut when SoFIA eventually transitions to Intel's own process. As for Rockchip, they were only behind one of the three SoFIA designs that launched. You can tell which one by the "RK" in its model number.

Intel's strategy seems to be to use Rockchip as a "flex" partner, to provide proliferation designs as needed to address certain segmentation, in order to keep internal R&D leaner. Intel will need their own internal expertise to provide the base design, but partners like Rockchip and Spreadtrum may provide additional designs, as needed.

By the way, you forget that SoFIA will be used in phones - a market that Intel pretty much hasn't played in, until now. This will be incremental opportunity, not just a shifting of costs in low-end tablets.

Several hundred million in REVS by the END of 2016? And this is TABLETS.

This isn't unreasonable. If we assume by this timeframe that contra-revenue is no longer required, and revenue coming in is "several hundred million" per quarter by the end of 2016, and they reduce costs a bit in R&D, they can begin to approach profitability.