Here to make my stand, with a chipset in my hand

Yesterday, I clicked “like” on a LinkedIn post with the title “TI Cuts 1,700 Jobs”. Today, I read the analysis and pulled out Social Distortion’s “Still Alive” for inspiration. I’ve been through this more than once. For them it’s not like-worthy, and I feel their sting.

The part of the post I liked was the comment: “This is good for embedded.”

Most of the stories on this have been focused on how TI was unable to make lasting headway in the mobile business, citing factors ranging from a lack of a baseband solution to overall delays in product. I’ll let you read some of the other viewpoints:
Information Week: TI Cuts 1,700 Jobs, Exits Mobile Chip Market

ZDNet: How TI’s move out of smartphones into servers highlights chip issues

TechCrunch: Texas Instruments Cuts 1,700 Jobs As It’s Driven Away From Mobile Chip Market By The Rise Of Custom Chipmaking

CNN Money: Texas Instruments to lay off 1,700

Some of that is good insight. On the other side, my prior optimism for TI in the mobile biz seven months removed is now proven somewhat misplaced.

The difficulties of selling merchant chipsets into a mobile device ecosystem mined with patents and IP licensing, software support, and carrier preferences – and finicky consumers – are now becoming evident. The winners are clearly Apple, Qualcomm, and Samsung, companies that have critical mass in those four areas. (But, the folks that are using the Samsung model as an example of how to win may want to fact check what chip is inside the GS3 in North America. It’s not Samsung.)

None of these outlets seem to be taking TI’s repositioning to embedded very seriously, which is understandable for the mobile press – the last strategy didn’t work, so why should we believe you when you tell us about this one?
.
While other investment houses downgraded TI, Stifel Nicolaus issued a spot-on comment as to why shifting to embedded is credible (source: Dallas Morning News, Sept. 26, apologies for extended quote but it is pertinent)
“TI’s new strategy will drive “higher quality revenue” in embedded processing to fuel growth and expand market share, Stifel analyst Tore Svanberg wrote today in a research report. TI’s OMAP and connectivity businesses have “struggled to become profitable and operated at a significant loss last quarter (and likely getting worse in the September quarter) and prospects for profitability remain challenged in the highly competitive and capital intensive smartphone and tablet markets,” he said.

“Svanberg noted that TI isn’t giving up entirely on wireless: It will continue to develop OMAP and connectivity solutions for the embedded market. Wireless connectivity is gaining momentum in the embedded market in smart appliances, automotive applications and mobile devices, he said.

“TI over the years has expanded its reach into embedded processing, which accounts for about 15 percent of total revenue (or about $2 billion). TI controls 12 percent of the $18 billion global market, holding the No. 2 spot.

“TI has made significant progress over the years to expand its product portfolio and we have confidence that TI can utilize its sales force and applications engineers (three to four times larger than its closest competitor), large product portfolio and unmatched customer base (90,000 existing customers and 60,000 untapped by embedded processing) to drive share gains and further customer engagements,” Svanberg said.”

Any time you can shift energy from a trailing and fading position in a huge market to a leading position in a big and growing market, do it, in spite of the short term pain that might cause.

In order to create this shift, there is very little TI has to do technology-wise. Same ARM cores, and DSP and mixed-signal IP. Same focus on Linux, Android, and RTOS platforms. Perhaps more importantly, better leverage of a highly skilled sales channel that has knowledge and access to thousands of design-win opportunities. There are definitely cost and volume differences, but TI knows how to manage those. We naturally turn our attention to the biggest, most complex chips out there, when in fact this new world of 50B connected devices is going to have a lot more mid-range parts.

This is the same painful decision Freescale made a few years ago, electing to stop chasing the top of the Power Architecture roadmap (aka the IBM PowerPC 620) in the name of a couple design wins competing with Intel, and instead focus on more, but smaller e500 and similar cores where Intel wasn’t (pre-Atom, but that’s another story). The result was more design wins and better business.

TI is now facing that same point, with a lot more than just OMAP in play. There are new families like the KeyStone DSP+ARM and the Hercules ARM Safety MCU targeting a range of embedded applications. Toshiba and RIM may have done TI a huge favor waffling on OMAP-based product launches, and the Amazon Kindle Fire HD will unwind naturally. OMAP chipsets will certainly be rebranded, but the combination of processor plus graphics plus connectivity and analog interfaces will fit in a lot of places.

It’s sad, but it is by no means over for TI.