Amazon Considering Buying Texas Instrument's Chip Business

puddingebola writes "From the article, "Amazon is reportedly in 'advanced negotiations' to acquire Texas Instruments' OMAP chip division, bringing chip design for its Kindle tablets in-house, and helping TI refocus on embedded systems. The deal in discussion, Calcalist reports, follows TI's public distancing from its own phone and tablet chip business in the face of rising competition from Qualcomm, Samsung, and others, though Amazon taking charge of OMAP could leave rivals Barnes & Noble in a tricky situation.'"

Comments

Big Move

[roidzrus]

This is a big move for Amazon, especially considering their Kindle doesn't necessarily demand huge amounts of performance, especially not in comparison to the iPad and other high-end Android tablets. I never really thought they would go this way, but now I can't help but wonder if they're going to expand toward phones as well. It seems that they could have just as easily sourced chips from Qualcomm unless they had something huge planned.

Re:Big Move

[vlm]

This reminds me of Motorola selling off all of their real industry (e.g. creation of ON Semiconductor) and turning into nothing more than a junk mobile phone company paying manufacturers to make the junk overseas. Motorola used to be the REAL DEAL.

In other words, copy HPs business model.

Re:Big Move

[SJHillman]

eBook readers are less of a fad and more of an intermediary step between books and tablets. I have a Nook Color, which further blurs the line between the two (it's more of a tablet optimized for reading books and magazines).

Re:Big Move

[rtb61]

The "Patent Wars" heat up and if you don't think all this and other recent manoeuvring is all about patents and Apple's triggering of the "Patent Wars", then you really aren't paying attention.

And the Beagleboard, Samsung Products, et al

[eldavojohn]

Amazon taking charge of OMAP could leave rivals Barnes & Noble in a tricky situation

Also, I believe the BeagleBoard is the SoC OMAP3530 ... not to mention there's a bunch of Samsung products (since it was mentioned that they are "rising competition") that depend on the OMAP4xxx series like the Galaxy S II and Galaxy Tab 2 and Galaxy Nexus ... lot of BlackBerry devices on that list too. It's not just the Kindle Fire using OMAP4, there's a lot of current devices using OMAP3 & OMAP4.

What's going to happen to all these devices when Amazon decides it doesn't make open source hobby boards or cell phones and condenses these SoCs down to just Kindle-related focus? I guess it'd be stupid to throw away all that business but anybody know what would happen to these?

Re:And the Beagleboard, Samsung Products, et al

[phantomfive]

Samsung, B&N, Blackberry etc will have no problem switching to another provider, probably Qualcomm. ARM chips from TI are fine, but they are definitely not the only provider, and switching from one to another is mainly a matter of changing drivers. It's not a big deal.

The hobbyists might have more trouble though, although the BeagleBoard is sold under the CC license, and in any case you could switch to arduino possibly, so they won't be completely lost.

Re:And the Beagleboard, Samsung Products, et al

[Svartalf]

Quite simply, there's enough range of stuff (including the Embedded Sitara based BeagleBone...hint...hint...) that you can lay hands on for that price range right now from alternate suppliers of A8/A9 based SoC's.

It's annoying, but you should be thinking in terms of being able to jump ship to something else on a moment's notice- they're in it for the money and they can drop anything anytime they feel like it. Even if TI doesn't sell this stuff...they could still end up pulling the plug on it all. It's just unlikely if TI's still doing the OMAP stuff- they use BeagleBoards and PandaBoards as internal dev boards for a lot of projects at TI. Now...if Amazon gets the division, it might be a differing story, maybe not- your guess is as good as mine. They've got to get dev boards for their own stuff and before they build out new models- it may be that nothing changes in the medium term because of that fact.

If you're concerned, the Allwinner based devices are a good prospect as is the Igloo which uses the Samsung SoC lineup instead. Worrying about whether or not you're going to lose the Beagleboard/Pandaboard is really kind of fruitless. Go elsewhere if you're concerned.

Re:And the Beagleboard, Samsung Products, et al

[Svartalf]

It should also be noted that if they sell the division, TI or it's successor in interest is OBLIGATED to fufill any obligations for supply lifetime on parts- period. For TI to have gotten the business in the first place, they'd have to guarantee it. It's how that part of the industry works.

I can assure you that the vendors of these devices will have 5-10 year windows on parts that WILL be honored or Amazon and TI would be on the receiving (read: LOSING) end of a nasty breach of agreement lawsuit.

What?

[Alwin Henseler]

Call me ignorant, but since when is Amazon a company that develops hardware?

I know Amazon has a big catalog, but customized / re-branded products aside, aren't they basically a box-moving company? What the *** are they doing in the chip development business? More specifically: what do they expect to do, that a specialist like TI can't do for them?

Re:What?

[taktoa]

http://www.eevblog.com/2012/10/15/eevblog-370-kindle-paperwhite-teardown-review/

They're pretty fucking good at hardware

Re:What?

[SpzToid]

Have you heard about the Amazon Kindle? About the only thing you can compare it to in terms of actual units sold is the iPad, so Amazon is in Apple's league. Yes, seriously, Amazon is huge in 'readers' which is a variant of tablets, I suppose as I don't have much experience with them myself.

Just as Google developed the Chrome web browser so as to have direct influence over the presentation of the web, Amazon has created hardware readers to have direct influence over how electronic versions of its content are consumed.

Neither wants to be at the mercy of some vendor, (and I'm thinking about you Adobe and your Macromedia Flash).

Re:What?

[Jeng]

problems include not being able to ship on date in ~10 years or having the promised performance

I did a few weeks of temp work for TI once. I'm not sure why they pulled us from shipping, but we were looking over circuit boards I believe for routers or something. Some of the chips were on upside down, as in shiny side up, and these has already passed two QC's. That is not what they were having us look for, but it is what I found. They no longer had a need for my services soon after.

Re:What?

[BBF_BBF]

Call me ignorant, but since when is Amazon a company that develops hardware?

I know Amazon has a big catalog, but customized / re-branded products aside, aren't they basically a box-moving company? What the *** are they doing in the chip development business? More specifically: what do they expect to do, that a specialist like TI can't do for them?

Umm... Amazon runs one of the largest commercial "cloud" computing services, they've moved from being a strictly "box mover" a LONG time ago. Buying TI's OMAP division may help Amazon develop custom low power servers that would improve their "cloud" in addition to moving their e-reader/multimedia consumption device design in-house.

Re:What?

[uvajed_ekil]

Call me ignorant, but since when is Amazon a company that develops hardware?

You've got it. Amazon don't develop hardware like this, which is why to get in on they act they need to buy an established player. Whether this is a good idea or not is open for debate (looks okay short term, but I'm highly skeptical beyond that), as it would bring Amazon something very different that is out of their wheelhouse. As a company that does services and supply very well, is it really necessary to go so far with the Kindle? The e-reader business is a money maker of a delivery system for them but the content is still king, so while developing hardware in-house could lead to a better or cheaper Kindle, it could also limit flexibility as competitors innovate, making this all a big gamble. Since the hardware and specs are less important to Amazon's business it would certainly be safer to continue to adapt what is already available and simply stay in the race with Apple et al., rather than trying to beat them at their own game or change the game entirely. If I were in charge at Amazon, this is problem not a problem I would throw a ton of money at, or at least not my number one solution.

Misleading Title

[Anonymous Coward]

OMAP is only one small part of TI's integrated circuit business.
 
That said, I'd really prefer if they kept it. I really like what TI has been doing with OMAP lately. I'm afraid Amazon might ruin it for the rest of us.

Amazon needs to focus

[Nimey]

Are they selling Kindles or books?

Right now they seem to be laboring under the delusion that they're a Kindle company. They've deliberately crippled the Android version of the Kindle software; for instance, you can't categorize books into collections, which is a Big Deal if you're a serious reader and have dozens or hundreds of books. There's no technical reason for the omission, it's just market segmentation.

I can't imagine they (currently) make more profit on the Kindle than they do selling their books, music, et al, so I can't see a rational reason for this strategy.

Whats the point of an intermediary step?

[Viol8]

Its not like you need to train on an ereader first before you buy a tablet. They're a fad , nothing more. In 10 years they'll be just another long forgotten footnote in tech history.

Re:Whats the point of an intermediary step?

[boristdog]

Until tablets have a hell of a lot better battery life, e-readers are not a fad. I can take my e-reader on a month-long trip, read a couple hours every day and still be reading on the same charge when I get back. I charge mine once every couple of months.

I think you miss the point that a lot of people use e-readers to read (a book replacement) rather than to surf the web or do e-mail (a computer replacement).

Re:Whats the point of an intermediary step?

[0123456]

Its not like you need to train on an ereader first before you buy a tablet. They're a fad , nothing more. In 10 years they'll be just another long forgotten footnote in tech history.

Tablets may be a fad, but they will still be here in ten years.

Oh, you meant e-readers? They exist because they're a fsck-load easier for most people to read on than a backlit LCD, and because losing a $60 e-ink Kindle when you leave it on your chair by the pool is much less disastrous than losing a $600 iPad where you stored all your login passwords.

Not only that, but before long e-ink e-readers will cost less than a hardback book. At that point they become pretty much disposable items.

Re:Whats the point of an intermediary step?

[Beorytis]

Also...

Until tablets have a hell of a lot better battery life,

and a display that's as comfortable as e-ink

e-readers are not a fad.

Re:Whats the point of an intermediary step?

[gman003]

The intermediary step is for the *technology*, not the *user*.

Dedicated eBooks have far, far better battery life, and are cheaper and often lighter. That means they work better in many of the use cases for the tech they're replacing, "paper books".

Tablets will eventually be able to encompass those features - they already do, for some people. Eventually. We're close enough that we can see the eReader is just a transitional phase, but it's a necessary stepping-stone.

Not getting into the chip *making* business

[robot256]

No one seems to have mentioned it yet, but it's worth pointing out that Amazon is presumably buying just the OMAP processor *design* unit, not the manufacturing unit. They will likely still use TI's foundries to make the parts, but Amazon will have control over the architecture and who gets the documentation.

Also worth reinforcing that this is not a bad deal for TI. ARM CPUs are pretty much a commodity product at this point, without much room for differentiation unless you go hog wild with optimizations like Qualcomm has. TI's main business has always been in the low-level ASIC and microcontroller markets, where is has a very large, well-respected variety of parts and continues to improve them.

Still waiting on a server version of OMAP

[Big_Breaker]

I really wish someone would come out with a dual core arm SoC with e-sata and gigabit ethernet as a light duty server. 1BG of ram would be nice but 512MB would probably work too. Guruplug, some of the allwinner media boxes and BYO drives NAS boxes come close but each miss something or cost too much. Single core is weak for rar / par and can get bogged down (yea, yea scheduler, blah blah). USB can't set HD parameters, has material cpu overhead and is wonky for RAID. 100/10 is a bit weak throughput for even a SOHO server.

OMAP5 with the right configuration would get there. Please stop putting 100/10 interfaces on these chips @&#$%!

I'm ok paying for a light-duty gpu and hdmi display interface but e-sata and gigabit pls, pls ,pls.

Re:Still waiting on a server version of OMAP

[Svartalf]

Marvell's done that...with up to quad cores even. Dell partnered up with them to make their first ARM server in their lineup.