It's true that individual SKUs have been EOLed and obsoleted since the dawn of the semiconductor age. I have RCA and Mullard databooks from the '60s that have notings like "Obsolete Part with Successor Part", etc. Planned Obsolescence has always been around.
The slight difference this time around is that several of the parts have no successor (LME498xx) and there are some parts where the suggested successor part is of lower spec than the obsoleted part (LME49990). Clearly, there has been some indiscriminate bean-counting at work.
I'm grateful that there are some single-sourced bipolar-only parts (LM3886, LM1875, LM6172, etc.) that have been retained, but it's unclear for how long. If this is a canary in a coal mine, I'd say that the prognosis doesn't look good for many of these surviving parts going forward. None of those has a viable alternative, either.
The slight difference this time around is that several of the parts have no successor (LME498xx) and there are some parts where the suggested successor part is of lower spec than the obsoleted part (LME49990). Clearly, there has been some indiscriminate bean-counting at work.
I'm grateful that there are some single-sourced bipolar-only parts (LM3886, LM1875, LM6172, etc.) that have been retained, but it's unclear for how long. If this is a canary in a coal mine, I'd say that the prognosis doesn't look good for many of these surviving parts going forward. None of those has a viable alternative, either.
I did some trolling through the last few years worth of Texas Instruments financial news. In the first paragraph of this article they have the opinion that T.I. tremendously overpaid for National back in the 2011 buyout. $6.5 billion, which was apparently a 75% premium to the National market cap at the time:
Dear TI, the Deal Clock Is Ticking: Brooke Sutherland - Bloomberg Business
Makes the situation even more curious. T.I. is essentially flushing a substantial part of $6.5 billion right down the loo here. Seems like the kind of thing some exec would lose a bonus over, at the least, lol.
Executive management at T.I. might not be too swift these days. In that article I see mention of a $7.5 billion share repurchase program. Seems that such programs have been roundly discredited as a ridiculous waste of capital in attempts to boost stock price, over using the money to simply grow the company. Kind of indicates the company is out of (growth) ideas.
Dear TI, the Deal Clock Is Ticking: Brooke Sutherland - Bloomberg Business
Makes the situation even more curious. T.I. is essentially flushing a substantial part of $6.5 billion right down the loo here. Seems like the kind of thing some exec would lose a bonus over, at the least, lol.
Executive management at T.I. might not be too swift these days. In that article I see mention of a $7.5 billion share repurchase program. Seems that such programs have been roundly discredited as a ridiculous waste of capital in attempts to boost stock price, over using the money to simply grow the company. Kind of indicates the company is out of (growth) ideas.
Last edited:
If these parts were a cash cow for the company, they would have either continued maintenance on the 6" line* or transferred the technolgy to another (8") line. Even the most vile bean-counter wouldn't overlook that.
*I admittedly have no idea what current offerings are for 6" tech at Applied Materials/Tencor.
*I admittedly have no idea what current offerings are for 6" tech at Applied Materials/Tencor.
I checked RS (UK) stock levels for the 4562 (8pin DIP) 12 hours ago. 399 in stock. This morning, ZERO. The soic outline is the same.
And it wasn't me lol.
(Farnell have hundreds though... at the moment
)
Like most things, this could be an opportunity to work with devices from other manufacturers and for new favourites to emerge.
And it wasn't me lol.
(Farnell have hundreds though... at the moment
)Like most things, this could be an opportunity to work with devices from other manufacturers and for new favourites to emerge.
I don't think anyone is building new machines for 150mm wafers. It's all retrofit at this point as 150mm plants get EOL'd--and probably moved over to III/V or other substrates, which I think tops out with 150mm wafers. I could be wrong, though (probably!).
Pretty much everything new is 300mm with a *little* bit of 200mm. Most 200mm is retrofit from fab consolidations, too. There might be a resurgence in 200mm for smaller lots on low power/larger node processes. Or weird things like high-performance analog.
I'm completely unsure how old this 150mm fab is (probably PRETTY OLD) and there may have been a plan even at National to close it down and move these chips to another line--even an old fab is going to have a lot of operating costs and the machinery doesn't last forever. After TI's acquisition, it'd be pretty easy to see a huge number of these chips being "close enough" to other designs in their catalogs that it wasn't worth redeveloping these chips on a new fab.
Pretty much everything new is 300mm with a *little* bit of 200mm. Most 200mm is retrofit from fab consolidations, too. There might be a resurgence in 200mm for smaller lots on low power/larger node processes. Or weird things like high-performance analog.
There are pockets of exotic stuff on small wafers, some starting material is very hard to make in large sizes.
The public has become numb to planned obsolescence or the fact that you don't take a dvd player to get repaired you throw it away. It's funny I just last week came across an industry with the longest qual time/life cycle that I have ever seen, flame monitoring for major foundry furnaces, failure is absolutely not an option.
Last edited:
Yes, Scott, sorry if I didn't make that clear. I think the biggest III/V stuff (and I'd hardly call that exotic nowadays) is 150mm? And much of it still on 100mm. I remember when a sapphire fab (defense communications stuff) came and gave a talk at my uni that they're on 150mm, too. I was honestly flabbergasted they had wafers that big for that substrate.
Last edited:
Space qual stuff used to be like that. Generally things would be EOL for everything on terra firma before it could get qualified for orbit. The UoSat projects have done great work showing you can do it with parts from mouser over the years.
Member
Joined 2009
Paid Member
The National Semiconductor stock traded at about $12-13 when news of the acquisition drove it to $24.9. The acquisition guaranteed $25 per share.
In the years leading up to the acquisition, National had been generating $1.5-2B in yearly revenue. That's 3-4 years to break-even on the $6.5B payment.
TI didn't buy National for their audio portfolio. I bet TI bought National for their superior analog high-speed precision BiCMOS processes, which are still very much alive and well.
Many companies across the US have been doing that. Companies are sitting on stockpiles of cash and use some of it to drive the share price up. Some refer to this as a liquidity trap.
Tom
Some timing details on TI Support forum.
Process ends September 28, 2016, but they will be shipping through March 28, 2017.
Process ends September 28, 2016, but they will be shipping through March 28, 2017.
I noticed the LMC567CN will be discontinued but not the LM567CN. Anyone know what’s going on there?
Revenue isn't profit, they would have needed many more years at that revenue to return a break-even unless they had no costs.
As far as the parts, probably the one I'm most bummed about is the LME49810/11/30 - sat in a niche that nobody else seems to be in for sure.
The consolidation of the profile was inevitable since the purchase of Natsemi. I wonder how many of the non-replaced parts will be licensed to second source brands - though probably none of the hero analogue parts as the know-how is still cutting edge.
Here are the replacement list given from the TI side..
LMxxxx EOL Information - Audio Amplifiers Forum - Audio Amplifiers - TI E2E Community
IMHO, all it's about cost & cost saving. A manager syndrome
and not competing against it's own opa's Many uses the LME49990 as the best ADC driver (see AKM ADC evaluation boards to get a better 10 dB THD+N), read also the running dyaudio project http://www.diyaudio.com/forums/equipment-tools/277808-diy-audio-analyzer-ak5397-ak5394a-ak4490.html...
Hp
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.