Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/intel-foundry-gathers-customers-and-partners-outlines-priorities-intel-connect-live.22714/page-6
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021770
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Intel Foundry Gathers Customers and Partners, Outlines Priorities (Intel Connect Live)

Intel's Arrow Lake processors, officially known as the Core Ultra 200 series, debuted on October 24, 2024, marking a significant advancement in Intel's desktop and mobile CPU offerings.

Is this the first Intel chip with TSMC N3? How well has it done?


View attachment 3148
I hate to say but it's very mid in gaming and latency sensitive workload due to their rather conservative D2D/NGU/Ring Clock the only good thing is the power efficiency but than again it's 2 node shrink so that's a given. It's quite good in Multi Threaded applications.
 
@Daniel Nenni @nghanayem

So if it take 4+ years from commitment by customer to production volume, then we can assume that any design/project wins this year will only be revenue in 2029. correct?

At the leading edge, from wafer agreement to HVM, 3-4 years for sure. From tape-out to HVM 2-3 years. Depends on the size of the chip but on the leading edge they are pretty big. Chiplets help that of course. The majority of the chip die is not leading edge. Intel is using TSMC N3, N5, and N6.

Remember, TSMC builds fabs based on customer commitments and how long does it take to build a new fab? 2-3 years or more depending on the location.

My guess was always 2030 for intel to make it to #2 in the foundry business. Maybe sooner if Samsung continues to do badly. The foundry business is a marathon not a sprint.
 
At the leading edge, from wafer agreement to HVM, 3-4 years for sure. From tape-out to HVM 2-3 years. Depends on the size of the chip but on the leading edge they are pretty big. Chiplets help that of course. The majority of the chip die is not leading edge. Intel is using TSMC N3, N5, and N6.

Remember, TSMC builds fabs based on customer commitments and how long does it take to build a new fab? 2-3 years or more depending on the location.

My guess was always 2030 for intel to make it to #2 in the foundry business. Maybe sooner if Samsung continues to do badly. The foundry business is a marathon not a sprint.
great input. Since Intel does not have many commitments so far in 2025, Looks like 2030 might be more like top 5 best case.

the marathon gets pretty long when you lose 6B+ every year and have to sell businesses to balance cash flow...
 
great input. Since Intel does not have many commitments so far in 2025, Looks like 2030 might be more like top 5 best case.
the marathon gets pretty long when you lose 6B+ every year and have to sell businesses to balance cash flow...

I think you will see some pretty big cuts at the end of Q2.

I also see more assets being sold off. Lip-Bu wants laser focus. Nvidia is a big Cadence customer, Lip-BU and Jensen Huang are speaking at the CDNLive event today. I think Lip-Bu will open a lot of doors for intel Foundry. As long as 18A and 14A are good enough business will come.

"The Intel-UMC partnership targets the competitive 12nm process node for various applications, with a projected market growth to $20 billion by 2028."
 
"The Intel-UMC partnership targets the competitive 12nm process node for various applications, with a projected market growth to $20 billion by 2028."
I love the UMC partnership but to be clear the TAM for ALL 12nm is expected to be $20B in 2028. the UMC+Intel goal is to have 5-10% share

Since we already established that it is 3-4 years from signing agreements to production revenue. The designs are already started on 12nm to have material revenue in 2028? correct? when was PDK1.0 delivered?

the revenue will show as UMC revenue with Intel getting leasing income?
 
I love the UMC partnership but to be clear the TAM for ALL 12nm is expected to be $20B in 2028. the UMC+Intel goal is to have 5-10% share

Since we already established that it is 3-4 years from signing agreements to production revenue. The designs are already started on 12nm to have material revenue in 2028? correct? when was PDK1.0 delivered?

the revenue will show as UMC revenue with Intel getting leasing income?
1746704113540.png

UMC/Intel 12 timeline.
 
Back
Top