Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/panther-lake-delay.22214/
        )

    [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
)

Panther lake delay?

Y.H

Member
Is this credible at all?

According to my latest industry survey of PC/NB vendors (brands, ODMs, and EMS), Intel has postponed the mass production timeline for Panther Lake (PTL) from early September 2025 to mid-Q4 2025. Several key points warrant attention:

1. Given the typical 2-4 week gap between chip and finished product (PC/NB) shipments, PTL notebooks may not become widely available until 2026, implying that Intel will miss the crucial year-end holiday sales season in 2025 and 2H25 revenue and profit will face downside risks.

2. While Intel has only publicly stated that PTL is scheduled for mass production in 2H25 (keeping this delay within their communicated timeline), such a decision would not be made without significant production issues. This delay will impact Intel's 2H25 revenue, profits, and, most importantly, organizational morale and supply chain trust.

3. Defect density is just one-factor affecting chip yield and mass production capability. The shipment schedules of PC/NB brands, ODMs, and EMS providers serve as critical indicators for assessing the production status of Intel's 18A process.

4. Based on my survey of Intel’s advanced-node consumable shipments, 3Q25 figures show no major changes from current levels, which could support the delay of Intel PTL.

5. In 2H25, Intel will primarily rely on Arrow Lake (ARL) to compete against AMD and Qualcomm. With ARL offering less than 40 TOPs and brands showing limited enthusiasm for Lunar Lake (LNL), Intel appears disadvantaged in the 2H25 AI PC competition.

6. More aggressive brands have planned to adopt the limited Qualification Sample (QS) PTL chips expected in late September for specific high-end models to gain a time-to-market advantage. However, given the limited QS shipments, Intel will benefit minimally.

7. Until Intel successfully ships its own chips manufactured on the 18A, it will struggle to gain the trust of external IC design customers and secure their substantial resource investments for 18A chip development collaboration.

 
Historically Intel makes decisions in February/March, then communicates them to OEMs in June in roadmap update, then sometimes publicly (they never announces the meteor lake delay).

I would expect them to follow this. Most people were expecting paper launch at best in 2025.

Our model is that Panther Lake will be VERY expensive due to process complexity and trying to ramp in new factory that is not fully loaded. Again, Intel has created financial problems that are not easy to fix.
 
Historically Intel makes decisions in February/March, then communicates them to OEMs in June in roadmap update, then sometimes publicly (they never announces the meteor lake delay).

I would expect them to follow this. Most people were expecting paper launch at best in 2025.

Our model is that Panther Lake will be VERY expensive due to process complexity and trying to ramp in new factory that is not fully loaded. Again, Intel has created financial problems that are not easy to fix.
it will be cheaper than arrow lake for their margins no?
 
That is not clear. 18A will have cost issues until it is fully ramped into a full capacity fab.
The first time ramp sure no doubt about it but long terms it's fine they need to just make product on IFS First and move away most of the silicon from external to Internal.
 
Is this credible at all?

According to my latest industry survey of PC/NB vendors (brands, ODMs, and EMS), Intel has postponed the mass production timeline for Panther Lake (PTL) from early September 2025 to mid-Q4 2025. Several key points warrant attention:

1. Given the typical 2-4 week gap between chip and finished product (PC/NB) shipments, PTL notebooks may not become widely available until 2026, implying that Intel will miss the crucial year-end holiday sales season in 2025 and 2H25 revenue and profit will face downside risks.

2. While Intel has only publicly stated that PTL is scheduled for mass production in 2H25 (keeping this delay within their communicated timeline), such a decision would not be made without significant production issues. This delay will impact Intel's 2H25 revenue, profits, and, most importantly, organizational morale and supply chain trust.

3. Defect density is just one-factor affecting chip yield and mass production capability. The shipment schedules of PC/NB brands, ODMs, and EMS providers serve as critical indicators for assessing the production status of Intel's 18A process.

4. Based on my survey of Intel’s advanced-node consumable shipments, 3Q25 figures show no major changes from current levels, which could support the delay of Intel PTL.

5. In 2H25, Intel will primarily rely on Arrow Lake (ARL) to compete against AMD and Qualcomm. With ARL offering less than 40 TOPs and brands showing limited enthusiasm for Lunar Lake (LNL), Intel appears disadvantaged in the 2H25 AI PC competition.

6. More aggressive brands have planned to adopt the limited Qualification Sample (QS) PTL chips expected in late September for specific high-end models to gain a time-to-market advantage. However, given the limited QS shipments, Intel will benefit minimally.

7. Until Intel successfully ships its own chips manufactured on the 18A, it will struggle to gain the trust of external IC design customers and secure their substantial resource investments for 18A chip development collaboration.

This report comes from the same guy who is claiming 18A has 20-30% yields. This feels to me like he is just doubling down on his previous speculative claims.

To the best of my knowledge Intel has not given a specific month for the launch date of 18A, so it is awfully hard to postpone something that hasn't had a date set yet.
 
This report comes from the same guy who is claiming 18A has 20-30% yields. This feels to me like he is just doubling down on his previous speculative claims.

To the best of my knowledge Intel has not given a specific month for the launch date of 18A, so it is awfully hard to postpone something that hasn't had a date set yet.
I agree they are saying H2 25 July-December time frame so a slip from September to November is still on schedule.
 
They need internal and 10B/year in external revenue to be successful. If you dont have both, you need to outsource. The realistic external foundry predictions are much smaller.
Previously DZ mentioned they would need small amount of external IFS volumes to make the foundry model work. I guess to break even for IFS, the external revenue requirement is much less than 10B/year.

If they can get 10B/year for external volumes, that would put them near the second position in the foundry business just based on external volumes.

 
Previously DZ mentioned they would need small amount of external IFS volumes to make the foundry model work. I guess to break even for IFS, the external revenue requirement is much less than 10B/year
To be fair that was before internal wafer requirements cratered for the second time. It was also before various initiatives to get more capital efficient (although some amount of that was probably assumed in DZ's model on incomplete data from the new accounting). After all Intel originally planned around internal buisness plus a small amount of external business to require 5 long term 18A HVM fabs. Now they are at 3. Presumably that means unless cost saving measures are wildly successful, I'd assume that Intel might actually need a more meaningful amount of external volume to get a reasonable RoIC. Rather than internal volumes alone being able to deliver this and external business being gravy on top.
 
Previously DZ mentioned they would need small amount of external IFS volumes to make the foundry model work. I guess to break even for IFS, the external revenue requirement is much less than 10B/year.

If they can get 10B/year for external volumes, that would put them near the second position in the foundry business just based on external volumes.

100% agree. the problem is that they are not on track. If Intel gets to be second place for foundry, they will be amazingly successful. The are currently not top 10 (intel publishes external foundry numbers)
 
100% agree. the problem is that they are not on track. If Intel gets to be second place for foundry, they will be amazingly successful. The are currently not top 10 (intel publishes external foundry numbers)
I think for them, their primary objective should be to get IFS break even asap, as once the limiting factor on the valuation is removed, the narratives will be changed. It can then focus on profitability for IFS.
 
Another possibility - they could be delaying to give more time to get old product out the door before launching Panther Lake. In the mobile space they already have Meteor Lake, Lunar Lake, Arrow Lake-H all in production.

The latter two likely have a commitment to a minimum number of wafers from TSMC, and MTL is probably going to be a delicate balance of trying to build and sell enough to minimize losses on Intel 4.

(Though if I to bet, I'd be "Heads" on 18A and/or Panther Lake itself needing a little more time to bake).
 
Panther lake 18A is to replace TSMC Arrow Lake and Lunar Lake.
The problem is that Intel may be forced to keep manufacturing Meteor Lake to make the old fab running.
 
This report comes from the same guy who is claiming 18A has 20-30% yields. This feels to me like he is just doubling down on his previous speculative claims.

To the best of my knowledge Intel has not given a specific month for the launch date of 18A, so it is awfully hard to postpone something that hasn't had a date set yet.
This guy is an analyst famous for supply chain intelligence from manufacturers, suppliers, and component makers. Thus his yield info is not so useful but he may have some insight regarding launch date from his cells in laptop or PC manufacturers.
 
Panther Lake vs Meteor Lake yield.png

To me, saying PTL yield is slightly ahead of MTL yield at a similar point in time during development, is not exactly the highest of praise.
 
Back
Top