Can you be more specific on this? In the US, the corporate tax is based on net taxable income. So if a business isn't turning a profit, it doesn't pay taxes in the US either.
Moor Insights: “[if] you spin out IFS before Design & IFS is healthy, it will fail”
- Thread starter Brady
- Start date
Pat and BoD went all in on IDM and invest in leading edge technology. That was the only way the fab side and continued silicon leadership would have been possible.
The only other alternative at that time was to sell off the fab side to TSMC or Samsung, government approval allowing. That was what Bob was exploring and why he got the boot. Of course licensing of TSMC and management of Intel at a fee was also possible but that would have been selling out Intel. Thus IDM2.0!
It’s debatable how good DCAI or CCG is compared to Apple, Qualcomm, AMD or Qualcomm. For decades they held process leadership and scale advantage and x86 and barely were competitive and many generations sorry!
Untill Intel achieves leadeship at 18A they won’t be a meaningful nor legitimate foundry alternative to TSMC.
If you really split the company CCG and DCAI has no reason to pick IFS. Why would you? Without CCG and DCAI as captive customers for 18A IFS has no business rational to exist nor continue. Sure they look good with Lunar but that is because they are on N3B. I doubt they are competitive with Apple nor stand ahead when Qualcomm and AMD show up on N3
No, only chance for Intel relevance and revival as product company relies on IFS success at 18A and 14A and provide cover for their marginal CCG and DCAI team. Togather they have scale and some business case to be able to scale on Foundry. Sadly a culture and leadership transformation and change is needed in IFS
Rapdius is a joke, IBM is irrelevant as a leader in real technology. Expect them to deliver a smelly turd of process to Japan
The only other alternative at that time was to sell off the fab side to TSMC or Samsung, government approval allowing. That was what Bob was exploring and why he got the boot. Of course licensing of TSMC and management of Intel at a fee was also possible but that would have been selling out Intel. Thus IDM2.0!
It’s debatable how good DCAI or CCG is compared to Apple, Qualcomm, AMD or Qualcomm. For decades they held process leadership and scale advantage and x86 and barely were competitive and many generations sorry!
Untill Intel achieves leadeship at 18A they won’t be a meaningful nor legitimate foundry alternative to TSMC.
If you really split the company CCG and DCAI has no reason to pick IFS. Why would you? Without CCG and DCAI as captive customers for 18A IFS has no business rational to exist nor continue. Sure they look good with Lunar but that is because they are on N3B. I doubt they are competitive with Apple nor stand ahead when Qualcomm and AMD show up on N3
No, only chance for Intel relevance and revival as product company relies on IFS success at 18A and 14A and provide cover for their marginal CCG and DCAI team. Togather they have scale and some business case to be able to scale on Foundry. Sadly a culture and leadership transformation and change is needed in IFS
Rapdius is a joke, IBM is irrelevant as a leader in real technology. Expect them to deliver a smelly turd of process to Japan
To paraphrase a Prime Video commercial I recently saw:
Prince Charming! Is that you? Your grammar isn't as good as I thought it would be!
-------------------------
Nonetheless, I agree with some of your points about Intel, especially about Pat and the BoD. He was their Prince Charming, and admitting he was a dubious choice would probably be too embarrassing until they are on the actual precipice of failure.
Artificer60
Well-known member
Japan provided support for their steel industry by importing machinery and equipment ,implemented tax exemption measures, exempted Imported machinery from import duties along with clauses that allowed for special depreciation in corporate tax. In addition, reductions were made to the municipal property taxes for participating enterprises. These policies reduced the tax burden on the profits the companies generated and allowed more of those profits to be reinvested in modernization and research.
hist78
Well-known member
"Pat and BoD went all in on IDM and invest in leading edge technology. That was the only way the fab side and continued silicon leadership would have been possible."
If the 'only way' that Intel and Pat Gelsinger have chosen is going to kill Intel, then it's not a viable or sensible path. It would be a way to commit suicide.
To begin with, Pat Gelsinger's IDM 2.0 business model is a non-starter. He is betting his entire company on proving that the semiconductor industry was wrong for the past 30+ years in adopting the fabless/foundry business model. Unfortunately, the fabless/foundry industry is too busy and has no interest in arguing against him. For Intel's competitors, the easiest and cheapest way to weaken Intel’s ability to compete is to encourage it to expand into the foundry business. Intel will spend billions to build foundries and an ecosystem that could ultimately end up as disposable toys for its competitors. Among the 12 companies actively engaged with Intel Foundry's 18A, as Pat Gelsinger stated, how many of them truly want to see a stronger or more successful Intel?
They don't want to see a Fab monopoly, either. Especially when TSMC is already raising prices in 2025.
Don't tell me that Samsung is a viable alternative,
hist78
Well-known member
They don't want to see a Fab monopoly, either. Especially when TSMC is already raising prices in 2025.
Don't tell me that Samsung is a viable alternative,
TSMC does not compete against those potential Intel Foundry customers, Intel does.
If you are the CEO of an Intel competitor, do you want to kill Intel first or you want to have a second supplier first?
Why can’t Samsung be a viable alternative? They have successfully done manufacturing in the US, have experience being a foundry and have a skilled workforce in South Korea. They are also a memory powerhouse and so it would be a mistake to discount them. Finally, they are a revered national champion. I know they have had some hiccups on recent nodes but nothing that they can’t recover from.They don't want to see a Fab monopoly, either. Especially when TSMC is already raising prices in 2025.
Don't tell me that Samsung is a viable alternative,
siliconbruh999
Well-known member
Yes but the cant stop a technology transfer between them japan is a us ally btw i think 18A would have been better bet than IBM 2nm
Let’s see in Fab18 you got a team watching and choosing who will be king!They don't want to see a Fab monopoly, either. Especially when TSMC is already raising prices in 2025.
Don't tell me that Samsung is a viable alternative,
If you pick IFS you don’t need to worry about neither CCG nor DCAI?
siliconbruh999
Well-known member
There are enough IFS customers who don't really view Intel as competitors, such as all the hyperscalers, arm based chip manufacturers, etc.
Even the current Intel management probably thought about splitting Intel into two truly independent companies in distant future, but it is not a great idea to do that while Intel is weak as of now.
Artificer60
Well-known member
Because as you said earlier "The broader point remains that there doesn’t seem to be much of a market for a foundry that struggles with delivering on time." Samsung falls into that category.
What % of leading edge design companies compete with Intel’s products? What % do not?
I think this will answer the viability of IDM in-house.
siliconbruh999
Well-known member
Many competes with intel some are Intel customer as well Google Microsoft Amazon to name a few
18A should be fine for chiplets. I see Intel possibly stumbling with HNA-EUV as they have set expectations again too high and the technology is not even close to being ready for HVM. TSMC is telling customers that and they are listening.
Samsung Foundry hiccups started at 28nm. 14nm was good then 10nm failed miserable. 7nm was okay but 5nm and 4nm did not yield and 3nm was an utter failure. Samsung ruled the NOT TSMC market which shrunk to under $1B at 3nm and Samsung 2nm is not looking too good right now. Intel will pick up some of the NOT TSMC market leaving Samsung with crumbs from the kids table, my opinion.
Thank you for this insight. Do you think Samsung will throw in the towel at the leading edge?
Not likely. The Korean ego driven semiconductor culture will not allow it as long as they dominate memory. Same as Intel for CPU only Intel is not dominating CPU anymore. The CPU market is changing and competition is heating up. In my opinion the US Government may save IFS as a security precaution through fear, uncertainty, and doubt about world peace etc…