Amazon in Talks to Buy Texas Instruments' Omaps CPU Division?
A few weeks ago TI announced plans to get out of supplying integrated system-on-a-chip components to the tablet and smartphone market, a move which would upset B&N, Amazon, and a number of othe gadget makers, all of whom are making devices built around an Omaps chip.
A new rumor out today suggests that Amazon didn’t take this news all that well, and According to the rumor they are in talks to buy the Omaps division:
Amazon is currently in advanced negotiations to acquire the mobile chips from Texas Instruments, and thus may compete with the major electronics manufacturers such as Qualcomm, Samsung, Intel, Anbidih, and Apple.
The deal would likely be worth billions, but I’m not so sure that one, the rumor is real, and two, whether Amazon actually has the funds available.
But it would make a lot of sense for Amazon. This would put them in a position to secure a core component for their hardware, and with luck they would be able to customize the chips to suit their own needs. If they did that then they would be following in the footsteps of Apple, which has been working with CPU makers to create one of a kind chips for iOS devices for the past several years.
This would also mark a big change for TI. I can recall from CES 2010 that they were making a big deal about how they could supply a custom SOC solution for ereaders which integrated the screen controller into the CPU. They even had a funky-looking but eye-catching demo unit which showed up at a couple trade shows.
Even if Amazon does strike a deal, I’m not so sure US regulators will let it go through. Amazon will control not only their own chip supplier but one which sells chips to Amazon’s competition. That’s going to give B&N, et al deep interest in trying to block it. I’m not sure they will succeed but they will try.
fjtorres October 15, 2012 um 10:46 am
Don’t fret about the regulators.
There is nothing terribly special about OMAP chips; they’re just one family in a flood of ARM-based SOC product lines. You’re familiar with most; Marvel, NVIDIA, Qualcomm, Apple, Samsung, etc. Hardly monopolistic and companies sell components to competitors all the time in mature businesses run by adults.
(Granted, that does not apply to publishing or B&N.) 😀
Amazon is bound by existing contracts so they won’t be able to cut the flow arbitrarily, nor will they be wanting to. Amazon delights in turning cost centers into profit centers and selling OMAP processors to B&N just let’s B&N subsidize Amazon’s next generation CPUs. 🙂
Of course, the moment the deal goes through B&N will hold a press conference to announce they will stop using OMAP processors in future products out of "high-minded principle".
But they’ll keep on quietly using them as long as they’re cheaper than the NVIDIA and Qualcomm alternatives. If anyone notices, they’ll deny it. (They can always tweak the firmware and erase the chip ID numbers.)
Nate Hoffelder October 15, 2012 um 1:47 pm
Yes. The fact that Samsung continues to sell components to Apple while they are fighting WWIIII never ceases to amaze me.
fjtorres October 15, 2012 um 8:30 pm
The reason they’re fighting is because they like to make as much money as possible as many ways as possible. That includes doing business with a competitor so if cooperating makes them money, they’ll cooperate. Once it stops making them money, they’ll move on.
Only foolish executives make it personal.
Nate Hoffelder October 18, 2012 um 7:57 am
That’s a marked contrast to book publishing and retail, where a large number of business decisions appear to be made out of spite. Take the books published by Amazon, for example. Lots of stores won’t stock them.
fjtorres October 18, 2012 um 9:17 am
That is why I added the adjective.
I think it’s a 19th century culture thing, in publishing.
Decisions based on spite invariably boomerang. And badly.
Leaving money on the table over "high principle" only makes you look like a hypocrite or an idiot afterwards. And there always is an afterwards.
Wise executives practice tech industry-style coopetition; where necessary, you cooperate, where possible you compete, and you always look for a way to make the other guy make money for you. The name of the game is making money, not thumbing your nose at somebody else.
In this case, if Amazon were to actually buy OMAP (discredited rumor or not) the wise move for B&N would be to negotiate CPU prices at cost or threaten to take their volume elsewhere, which would raise Amazon’s own costs. What their current managent would do is switch to Marvell or a chinese product even at a higher component cost.
Of course, Nook Media may be run differently once the spinoff is complete but I don’t expect the tech culture to override the publishing culture there any time soon.
fjtorres October 15, 2012 um 10:50 am
Actually, the important question nobody is asking is what is Amazon buying?
The existing designs and contracts?
The design and engineering staff?
IP? Perhaps DSP and/or Telecom patents?
If Amazon is *really* going to do a cellphone line they’ll need some telecom patents in their portfolio and if TI is getting out of the phone and tablet business they can get good cash out of the relevant patents.
Amazon buying those patents buys them protection from the entrenched phone vendors so the price might by a bargain. Plus they keep them out of competitors' hands.
flyingtoastr October 15, 2012 um 4:42 pm
Unless they’re planning on issuing tens of millions of shares of new stock (presuming TI even wants Amazon stock, which, given its massive P/E ratio, I’m not sure they would) I’m not sure how Amazon could afford this. Amazon is forecasting a negative quarter, and their actual cash-on-hand is pitifully small, it means they would probably have to take out a fairly substantial amount of debt to pay for it.
And, given the bubble their stock is riding on right now, dropping a huge debt bomb on the company for potentially little results (OMAP4 processors are on the way out) I don’t see it happening.
Then again, Samsung has proven that vertical integration is the way to make money in the tech sector, so maybe Bezos will try this out.
fjtorres October 15, 2012 um 8:32 pm
It would be if the OMAP4 processors were *all* they get out of the deal.
But *that* wouldn’t be worth "billions" as the report quotes. That is why I’m wondering what sweeteners TI is adding to the mix.
flyingtoastr October 15, 2012 um 10:10 pm
I think the processor production line, expertise, and current clients are worth the billions of dollars.
Again, the problem comes down to the fact that OMAP4 processors are getting dated. The 4470 is by all measures a very good SoC, but compared to the newer designs from other companies (like Qualcomm’s new S4 Krait Snapdragons or Samsung’s Exynos) it’s rather long-in-the-tooth. TI looks to be putting the development of the OMAP5 on hold as they wind down chipset production on top of that.
Which means Amazon will be starting without a current-generation processor to sell to other companies, having to potentially design one from scratch (or partially designed if they decide to continue with OMAP5 development). Meanwhile they’ll still have to invest heavily in new infrastructure to get competitive – for instance, TI’s current chips are all built at 45nm as opposed to the 32nm or 22nm that other high-end SoC’s are now using. All in all Amazon Processors™ wouldn’t be looking at profitability for a couple years at least, not a good sign for a company with as razor-thin margins and P/E bloat as Amazon.
Which goes back to the root of the problem – for the price is it worth loading up potentially billions of dollars in debt for Amazon? My hunch is that Bezos would like to, but this is probably something that shareholders would attempt to block. It’s too large of a cost for too little returns.
fjtorres October 16, 2012 um 7:54 am
Do you have insider info of *exactly* what they are buying?
Or another, more detailed source?
'Cause all the reports I’ve seen on the web are the same sketchy rumor.
And it’s just a rumor, interesting to consider but otherwise meaningless until confirmed by an actual announcement.
And assuming that Amazon buying OMAP means they are buying foundries is a bit of a reach; the product and the facilities are not the same thing nor inextricably tied. And, even if they were included in the deal, Amazon could just as easily sell off the foundries and keep the IP and the design staff.
Barring actual reliable data on what is in the bundle I’ll reserve judgment on the wisdom of the deal.
Peter October 16, 2012 um 3:59 pm
"Barring actual reliable data on what is in the bundle I’ll reserve judgment on the wisdom of the deal."
It doesn’t sound like he’s making a judgment on the wisdom of the deal- he’s making a judgment on the plausibility of the deal. Amazon just plain can’t afford it.
Look at the math-
Last quarter Amazon had $830 million in operating capital on the books.
Last week, they announced they are spending $1.2 billion to buy their corporate offices in this quarter and the next.
They’ve already reported they won’t make any money this quarter, so the money won’t be coming from operations between then and now.
Bottom line- Amazon is tapped out for now. They must have some plan to find the money to continue operations and already announced expenditures, but large acquisitions aren’t too likely in the near future.
Lyman October 17, 2012 um 4:16 am
Although TI said they were de-emphasizing smartphones that is a big leap from dumping OMAP. There objective was expand OMAP usage, not shrink it or selling it off. Smartphones and Tablets aren’t the dominate deployment of ARM based SoCs in terms of numbers. The more expensive ones, yes. In terms of total numbers, no.
The other issue is that the day after the Rueters report TI basically corrected this "conclusion" made by the analysts.
"… However, TI remains committed to the OMAP platform and its customer …. Though TI is accelerating the expansion of OMAP processors into a broader set of embedded applications, the team remains dedicated to advancing current mobile customers’ OMAP processor-based product lines …"
OMAP 5 is sampling
Rather than 'buy' OMAP it is far more likely that Amazon is working with them on a better mix of functionality they’d want in a future OMAP. That may require Amazon more directly pre-paying for the R&D to happen. TI is looking for OMAP growth outside of smartphones and general usage Android tablets. I think it is a leap to put e-reader into the general Android tablet segment. Even the LCD devices that are specialized tablets don’t necessarily need a completely custom SoC (or OMAP if it doesn’t future needs; e.g., needs widespread cellular modem built into the SoC. ).
An Amazon smartphone is about as loopy as buying OMAP. It makes little sense given Amazon’s position. An iPod Touch competitor perhaps ( if Amazon’s music+video+media gets more traction).