Intel previously announced that its advanced 18A process had entered risk production, with some customers planning to use the new node for test chip production. However, according to a report from Reuters, Intel CFO David Zinsner revealed that some of those customers later withdrew after trial production, leading to a lower-than-expected volume of committed demand.
25 replies (most recent on top)
For someone to consider 18A:
- Do we have existing supplier relationship - No
- Is the supplier proven in their field (Foundry) - No
- Is the product proven itself to provide a strategic advantage - No
All red flags.
High volume foundry customers are wary of risks of moving to a new supplier, even with a proven process. With a new process, it will take an impossible amount of convincing.
Low volume customers are generally not interested in cutting edge process nodes as they are cash strapped and highly cost conscious. Mature nodes give them lower cost.
For Intel 18A, only real hope is in getting Pather Lake into the market and let the process mature for 2 years.
There will be occassional cash rich low volume customers like MSFT who will risk it. But those ones won't be enough to make Foundry sustainable.
Stock will be worth toilet paper until the Foundry makes money. If the Foundry doesn't make money,Intel will go bankrupt. Stock back below $20 in a week or so.
The CFO told the fact. He did a great job.
DZ is a fraud! His punk-asss needs to see a prison shower!!!
I don't believe OP understands semiconductor manufacturing.
Lies, lies all lies!! Shitel has turned into a marketing junk company surviving on government bail money
The CFO has never been that frank before but suddenly OP wishes to believe him?
Sad, isn't it?
CFO knows it's not a good idea to lie anymore. He already got caught once.
Paper can not hide fire. The CFO knows it is better to release the fact.
Maybe if Intel could fab a whole chip (Panther and/or Lake or Nova Lake) in our own fabs there would be more customers. Nothing like trying to sell an Idea that we're not even fully committed to. Dum--ss!
If I didn't know better dz is trying to reduce stock price
I finally found out what the A in 18A stands for.
AintNobodyBuyingIt
I think DZ is trying with a considerable effort to quash the widely spread rumor of a big customer win, likely because that leak upset the customer involved.
That aside, the stock in the upper teens and low 20s is discounting any external customers helping to remove the excess capacity.
Kind of amusing that DZ is taken as being forthright when most of the time what he says is not taken that way. This time is not different.
note the limited impact on the stock price
You mean the move from $60 to $20?
Yeah, the market never believed anything coming out of Intel.
old news
note the limited impact on the stock price
@ag+1jv929dn4 Shut down the fabs for the failures of the folks designing trash chips? If you shutdown the fabs Intel will cease to exist.
Repeat same mistakes every node and expect different results, and its doing same in 14a as well . It’s d-mbest at the best try to innovate and deliver immature nodes and put all unnecessary RnD that customers don’t want .
Have any of the companies who did trials and then backed out released any details on why they made the decision? Was it problems with the 18A process or just the abysmal experience of working with Intel?
PG openly stated that he bet the company on 18A. The company can't continue to have a 2-5 year recovery window that never attains success. The penalties for losing this bet will likely be severe.
Dear world. We concede this battle but we remain fully committed to pushing out various timelines the future. Thank you and good day.
If 18A cannot make it, Intel should stop future development and cut the loss on its fabs. It is becoming more than a joke.
The risk production is for Panther Lake on 18A. Intel need to prove at least one product can go production. 18A is not proven on any product yet.
For external customers, they usually try test chips to evaluate if 18A works for them. This is like a test drive before deciding whether to buy the car. This test takes many months or even years. Try a test chip does not mean customer is committed.
Intel need to prove 18A is usable for its own product.
Intel lacks the nature of being a foundry: no foundry can only do one technology. The basic nature for foundry is you have the choice of multiple process technologies for customers.
Gee.. what a surprise!!
Yeah, well, make stupid mistakes win stupid prizes.