NXP nears deal for Freescale Semiconductor

Status
Not open for further replies.
Now confirmed, subject to approval but I don't see a problem with that.

I'll ask the obvious question, what will happen to Freescale and NXP MCU range?

We had reps from Freescale in recently, they were eagerly promoting the Kinetis range, of course they didn't mention anything about a merger. They did promise a 10 year "lifetime guarantee" on certain chips, I wonder if that will be honored by NXP.
 
Good questions indeed.

My guess is even people fairly high up within Freescale probably don't know the answers to such questions, at least at this early stage.
 
The cost of maintaining the current supply / range of chips is likely pretty low. The resources required are a sunk fixed cost that the acquirer is likely to milk as long as possible. Look at how long Dallas / Maxim made their 1-wire counter chips available after they were officially retired.

IMO the bigger issue is transitioning to a similar yet different ARM architecture ie having to deal with a familiar space but different errata and / or idiosyncrasies that are vendor specific. All of that is over my head!
 
Constantin, the problem likely would be yes, they can stockpile a supply of the expected chips, but once the fab that produced them has been retooled and/or closed, it would be impossible to make any additional chips. Even if the fab is still available for the process used for the chips, it may not economical to do a run to produce more.
 
Michael, thanks for the update. I was under the impression that unless the fab made a transition to a new die size / doping method, etc. that a manufacturer could simply let the photo/laser/x-Ray lithography sheets stay in storage and re activate them as needed. I'd like to think that the folk at NXP would not be silly enough to kill off a existing product line for the fun of it. After all, revenue from existing customers is one of the prime motivators for the sale in the first place.

Once you force a significant platform change, you risk customers defecting to rivals as there would be no delta re switching costs. Eliminating a rival isn't a good enough reason either. There are too many arm core manufacturing companies out there and the price of admission for newcomers is relatively low. My guess for the motivation re this acquisition is scale, ie being able to depreciate the significant investments across a wider volume of product. That indicates a positive incentive to maintain the existing stock of processors.
 
Constantin: That assumes somewhere in the world there is a fab that is still capable of doing the same process. But when there aren't enough orders to run the fab, the owners will either close it down or retool it to run something else. Fabs are very expensive to run (and build). I just tend to be skeptical that when a part is at the end of life, that sooner or later there won't be any more supply, no matter what was promised by the vendor.
 
So Freescale was acquired by NXP. Then NXP was acquired by Qualcomm. Now Qualcomm is to be acquired by Broadcom. If acquirements lead to acquirements, where will the chain of acquirements end? :rolleyes:

If the regulators are worth a damn at all they should block a Broadcom / Qualcomm deal. Putting those two together (with all of their sub-acquisitions) consolidates a large amount of wireless chip IP into a single company. Just a day or two ago Hock Tan was in the news for moving Broadcom's HQ back to the US, for the now very obvious reason of this Qualcomm deal (so they aren't considered "foreign" as though where a company's HQ is located is anything but an accounting gimmick). As soon as he has monopolized all the big players HQ will probably move again to wherever the biggest tax haven is.

It's more than just consolidation of suppliers. These acquisitions are gutting the EE chip design field, as every merger comes with a reduction in the engineering force. I've been on both sides of this, both merged in and out. Hock has no interest in maintaining a large R&D dept, so to get new products he needs to feed off new acquisitions. Ultimately the cycle degrades the overall system. There is now a situation where there are vast tracts of IP that the original designers are gone and whoever is there is left scratching their heads trying to figure it all out (similar to NASA's knowledge problem where all the guys who designed the moon landers retired or left). A couple decades ago this was a good field to be in, now it looks like doing grunt work in a startup (and hoping for a buyout), or working for the half-dozen remaining big corps (doing same grunt work). The software field is starting to look much better and diverse (and way easier to do a startup in - hardware startup doing chip IP is ridiculously expensive).
 
Looks like the right thing happened:
... executive order Monday blocking Broadcom Ltd. from pursuing its hostile takeover of Qualcomm Inc., scuttling a $117 billion deal that had been scrutinized by a secretive panel over the tie-up’s threat to U.S. national security.

Trump acted on a recommendation by the Committee on Foreign Investment in the U.S., which reviews acquisitions of American firms by foreign investors. The decision was unveiled just hours after Hock Tan, the chief executive officer of Singapore-based Broadcom, met with officials at the Pentagon in a last-ditch effort to salvage what would have been the biggest technology deal in history.

"There is credible evidence that leads me to believe that Broadcom Ltd.," by acquiring Qualcomm, "might take action that threatens to impair the national security of the United States," Trump said in the order released Monday evening in Washington.
 
Status
Not open for further replies.
Back
Top