Intel may sell part of Intel Foundry in the future - Intel at Citi 2025 Global TMT Conference
- Thread starter hist78
- Start date
MKWVentures
Moderator
What specifically do you not understand? Intel has released plenty of new products. The finances of these are interesting
siliconbruh999
Well-known member
I think he meant good products and in recent years only LNL/GNR fits in that category ARL-H/HX is fine in Mobile they haven't released a solid product portfolio since Alder Lake.
Intel will do more than $50B is revenue this year. AMD is still $20B behind Intel. I would not bet against Lip-Bu Tan. He has not been CEO for six months yet and look at the change in Intel. Give him another 18 months and you will see what I have been talking about, absolutely.
siliconbruh999
Well-known member
Intel puts more Capex and R&D though so there is that AMD just don't take that big of risk in terms of Volume
In terms of overall business scale, AMD is unlikely to surpass Intel before 2028—unless its AI-related product lines gain significant traction. However, AMD’s revenue market share in server, desktop, and laptop CPUs is consistently higher than its unit share, indicating that AMD is selling at higher prices—likely with better margins—compared to Intel.
R&D has never been a problem at Intel. Getting the right products out in the right timeframe is the Intel problem. As I have mentioned before, the massive Intel ego destroyed the company. That is gone with LBT.
Intel spent $16.5B on R&D in 2024, AMD spent $6.B in 2024. I'm sure you will see less R&D spending from Intel in the coming quarters but hopefully it will result in much better products. Unfortunately this transition will take time. Given the investments by the USG and Softbank LBT has a plan that is bankable. I just hope the Intel BoD does not get in his way.
siliconbruh999
Well-known member
I would like Intel do R&D spending in useful stuff not stuff that is canned 90% of the time
hist78
Well-known member
Intel spent $16.046 billion on R&D in 2024, which was 31.2% of its revenue in 2024. That’s a huge amount of money but still probably far from enough for Intel as an IDM.
Unlike Intel, most of its competitors (excluding Samsung, also an IDM) and Apple are focused on just one side: either manufacturing (foundries) or product design (fabless). Intel, however, has to spread its R&D budget between both. Once that $16 billion R&D gets divided across multiple manufacturing and product design divisions, it almost guarantees that Intel will only win a few battles but lose many more golden opportunities that competitors can fully capture.
siliconbruh999
Well-known member
Look at the Table Intel R&D is more than AMD+TSMC Combined they should be able to compete with TSMC+AMD combined they are not effectively utilizing their R&D.
But Apple also conducts R&D in a wide variety of fields, including processors, so we can't really criticize others, right?
The most important thing is not to narrow down the target, but to get the effect of R&D.
Is this the reason you're selling Fab?
In my opinion, Apple didn't make a smart choice either, as they have a habit of making everything in-house...
Such things always fail, just like Intel.
hist78
Well-known member
Yes, efficiency is a big problem for Intel’s R&D. In the semiconductor industry, when Intel spreads its R&D across both manufacturing and product design, the efficiency and results are unlikely to be great. Intel Product and Intel Foundry often have different, even competing, priorities and business goals.
On top of that, Intel may also suffer from poor R&D effectiveness because of its global locations. It has about 5,000 employees in Ireland and another 9,000 in Israel, covering both manufacturing and research. But Ireland has some of the highest living costs in Europe, and Israel is among the most expensive countries to live in the OECD.
Meanwhile, Intel has little to no R&D or manufacturing presence in South Korea, Japan, or Taiwan, three countries with world class strengths in semiconductor research and manufacturing. Intel has largely ignored them for years. As a result, it hasn’t developed fabs or R&D centers in more cost effective regions that could help lower its overall cost structure.
I also can’t recall ever buying consumer electronics that originated from Ireland or Israel. Compared with Taiwan, Japan, and South Korea, the broader semiconductor ecosystems in Ireland and Israel are much weaker.
Source: Google search.
MKWVentures
Moderator
For obvious microeconomic reasons, Manufacturing and R&D are done in different places by all companies.
In the US, much/most R&D is done in Bay Area.....
Intel has enormous manufacturing sites in Vietnam and China and Malaysia
hist78
Well-known member
Those sites are only for low margin test and assembly of Intel’s chips, not for fabs. Intel builds its fabs and R&D centers mostly in high cost, expensive locations and it’s killing the company.
MKWVentures
Moderator
So Intel should not build fabs in the US, correct? The total cost for Ireland fabs was cheaper than Taiwan option or Japan Option
But you agree that R&D is frequently not done in low cost Geos. Correct?
If Intel breaks out IFS like they say they will (all accounting separate), it will all become clear.
hist78
Well-known member
I didn’t say Intel shouldn’t build fabs in the US. The US is Intel’s home base, and it should make full use of it, just like South Korea does with Samsung and Taiwan with TSMC.
But Intel is a global company, and it needs to establish international presences to tap into diverse resources and stay competitive. Instead, Intel had chosen only expensive international locations for its research and fab sites outside the US. Does that make any sense?
Why has Intel ignored Japan, Taiwan, and South Korea altogether?
Last edited:
MKWVentures
Moderator
Intel looked at sites in Taiwan and Japan. In each case, Intel made what it thought was the most cost effective decision. and In fact, the newest site for a Fab that intel created was in Dalian China. New greenfield sites add a ton of cost.
Ireland was per EU requirement many years ago but the subsidies make it work and they have been a productive site.
As we will soon see with TSMC vs Intel in Arizona, the issue isnt location for Fabs, Its the process and the culture.
Intel had a another "quirk" that led to some site choices. Intel thinks they are the leader in all technologies. So they don't want to go to a site where people could be recruited away ("everyone wants to steal Intel people"). Its kind of Ironic now as Intel will get people from TSMC to help them become competitive in manufacturing.
hist78
Well-known member
Compared with Intel, Apple focuses solely on product design and doesn’t handle its own manufacturing. Intel, on the other hand, must manage both product design and manufacturing under its logic semiconductor IDM model. This makes Intel’s business far more complicated, extremely costly, and often full of conflicts of interest within the semiconductor industry.
Another way to look at it is by comparing Apple’s net profit margin not only with Intel, but also with Intel’s OEM customers such as Dell, HP, Lenovo, Asus, and Acer. In many market segments, these companies are Apple’s real competitors. The challenge is that Intel's OEM partners often struggle to generate healthy profits using Intel’s processors, while Apple enjoys far higher margins by focusing on design and outsourcing its manufacturing.
Last edited:
hist78
Well-known member
I agree. I just don’t like it when someone duplicates Intel’s entire 35-minute content onto their own YouTube channel and encourages everyone to watch it, even though the same Intel/Citi audio is freely available on the official site. This duplication is essentially piracy. Regardless of the tool used to copy it, this practice is harmful to the academic environment, society, and the semiconductor industry.