https://www.bgr.com/1921034/apple-intel-manufacture-future-m-series-chips/
20 replies (most recent on top)
It's Fruit of the Loom
Step 1 was to Steal Underwear. Mission Accomplished!
Step 2 was undefined, but getting some external customers will work.
Step 3: Profits!
Spinning off NEX and focusing AI on the market segments not dominated by NVDA both seem like steps towards gaining NVDA as a customer.
I'm sure NVDA, AAPL and all the other large tech companies have told the Board that they want to see IFS be spun off prior to signing contracts, as much as they would like to see a second source to TSMC survive.
Try to see what LBT and the Board are doing through the lens of gaining meaningful foundry customers, because getting IFS profitable is the single most important issue they need to solve. IFS is rapidly deteriorating the financials and that must be stopped.
I agree that potentially radical changes are coming to the fabs and that includes selling a number of them to another foundry. I think that is why the company was previously discussing bringing GF Ex-CEO Caufield on board, to manage the transition of at least some of the fabs to GF.
Likewise all those rumors of various product groups being looked at by various companies likely showed where the fire-sale valuation was at, and the Board decided it would be better to push product groups into the market instead, starting with NEX.
This stuff is all connected but never communicated, because that would reveal strategy.
Look towards a day when some more product groups have been spun out and fabs have been sold. Expect IFS to be spun out as well, once it has cleaned up the financials.
Removing all points of competition is the only way Intel can be partners with the major tech companies, and they need that foundry business because x86 is on a slow downward slope (and has been for at least 5 years).
He's leaning up foundry business for a full sale. If it fails. Oh boy.
Any customer comes to try the test chip is just a try, just like you have to take a look at the fruit (if it is good) before buy it. I am afraid because 14A is not ready and no hope to be ready in visible future, customers have all turned away.
@en The break from the excess fabs, tools and thus people is one of the biggest changes LBT has announced. He stated it clearly and the way the ongoing layoffs were focused on the fabs shows he means it.
Not sure how far back the excess capacity strategy goes but Andy Bryant repeatedly stated the virtue of always having a shell ready.
Maintaining excess capacity is a remnant of having market domination during the 1990s and early 2000s, with the (questionable) idea being that the company needed to be able to fill every order or risk losing that dominance.
That should have been stopped in the 2010s but lived on and has resulted in too many fabs, each of which has too many tools and therefore requires too many techs to maintain.
In IDM 1.0, cost was not given much consideration, with all the focus on being leading edge, and the yield ramp. IDM2.0 if anything made the situation worse, till it resulted in the longest consecutive stretch of quarterly company losses since it started making CPUs in the late 1980s.
Fixing this will require selling fabs and many even selling some of the tools in the existing fabs, or warehousing or cold-bagging them. Shutting down the tools reduces the number of fab techs needed to run the fab and is a fairly quick way to align capacity to demand in the short term.
IFS needs a complete restructuring if it is to be spun off or otherwise exist in the longer term, and that means getting small enough to be sustainable. Gaining external customers may shortcut that effort but it does seem like they won't show till 14A and the pressure to fix this can't wait that long.
So anyone working in the fab should know that massive change is coming and will change the fabs to be much more like working at Global foundries (or their fab may even be sold to GF).
The company has stated in various ways that the external volume needed is not that great. It is merely the difference between the volumes Intel currently has and what it had a few years ago, because the fabs used to not lose money. I don't have the exact number but it is probably out there on the internet somewhere.
They hope to pick up several customers and have had several meaningful customers checking out 18A, but the customers want a standard PDK, which means industry standard design rules, and they also really want IFS to be independent from Intel.
18A-P gets them a standard PDK and sometime between then and 14A I expect IFS to be spun off.
Difficult to spin it off while it continues to lose so much much money, but the company is now focused on rapidly reducing the excess capacity, and shedding the older fab sites seems inevitable at this point.
IFS would start up with 14A so anything older than that may not even be spun off, but rather it would be sold, or shuttered when the current roadmap runs out.
Intel fabs are non-standard, so selling them to one of the other foundries makes more sense while they are still busy, so the new owner can recoup some of the cost by selling wafers to Intel, then go through a retrofit once the Intel business is gone.
Apple is a whale, but M chips aren’t enough volume to justify 14A.
Bye bye LTD
14A is not in any shape to start test chip for customers to try. It is not developed yet, and would take 5+ years for TD development.
CEO said it would need customer to continue with 14A development. This is unrealistic. No customer will pay for something far away in 2030+.
Chapter 11 reorg then Chapter 13 liquidation ! Intel needed this mid 2000s ! Intel suffers from over bloated PPE and no process lines but only HP Dell and Lenovo ! Stop the BS construction projects ! Build, tear down redemo then rebuild is billions in wasted money gone in smoke running up Intel boards dreams !
Fire the Intel board !
@a1 fake news. AMD did not have leadership in the beginning but they did not draw any exit paper for chapter 11 lol
nothing is confirmed
Don't be fooled.
Apple is just using this to gain better pricing with TSMC.
If Intel would stop shooting themselves in the foot, they would already have a whale.
They just cant stop fu--ing up. I mean, stupid-a-s amateur level mistakes.
Other companies have looked at Intel IFS and then walked away. BTW when someone says they are considering something it does not mean it is a done deal so I am not sure how you see this as "secured".
It's Atari
To justify 14A expense you would need multiple whale customers lined up with massive orders.
@a1 spot a racist
@a1 you have a point. We have either layed off, or let most of the good engineers leave on there own. There really is not very many engineers left who have built anything for Intel. It is not that the workforce is bad, it is that they have never been part of anything good at Intel. Also, there is no new talent coming in…..it is a total cluster.
Right now the lawyers are drawing up exit papers for Chapter 11 on Intel corporation. This is the future of Intel. Listen once you lose the leadership in high tech..you are GONE! NO RECOVERY, NO WHALE, NO MORE COFFEE OR FRUIT...GET A CLUE AND READ THE ROOM.. ..
Answer me this OP...what group of engineers that are left at Intel could possibly have a clue about making anything competitive? The white guy with the pocket protector that could spew formulas off the top of his head have been replaced by ID--TS...GO IT??????