Intel has limited customer commitments for latest chip manufacturing tech, CFO say

[hist78]

"May 13 (Reuters) - The volume of processors Intel (INTC.O), is set to produce for external customers using its upcoming manufacturing technology is currently "not significant", finance chief David Zinsner said on Tuesday.

Committed volumes, or the amount of external customers' chips set to be produced by Intel using upcoming manufacturing tech, is presently not significant, Zinsner said at J.P. Morgan's Global Technology, Media and Communications conference being held in Boston, Massachusetts.

"We get test chips, and then some customers fall out of the test chips... So committed volume is not significant right now, for sure," Zinsner said."

https://www.reuters.com/business/intel-has-limited-customer-commitments-latest-chip-manufacturing-tech-cfo-says-2025-05-13/

 

[MKWVentures]

I always liked DZ... he keeps it real.

He is 100% correct.

 

[XYang2023]

They also do not need many customers.

 

[fansink]

"We get test chips, and then some customers fall out of the test chips... So committed volume is not significant right now, for sure," Zinsner said."

"So committed volume is not significant right now, for sure"

Code for....

7 short horn blasts followed by 1 long blast

Then, "All passengers and crew proceed to your assigned muster stations...This is not a drill"

 

[Xebec]

Did Microsoft back out of using 18A?

 

[fansink]

Did Microsoft back out of using 18A?

Everyone knew it would've been a token order at best.

If Microsoft is out, I'd put the fabless clock at 23:57

 

[hist78]

Did Microsoft back out of using 18A?

In terms of volume, potential server or AI accelerator orders placed by Microsoft or Amazon at Intel Foundry likely won’t be significant.

Assume Microsoft needs Intel Foundry to manufacture one million units per year of a particular in-house designed server CPU, and assume a 12-inch wafer yields 300 units of that CPU:


1,000,000 CPU units / 300 (units per wafer) = 3,333 wafers


3,333 wafers / 12 months = ~278 wafers per month


You can plug in your own assumptions, but you'll probably reach the same conclusion: the volume is not significant.

 

[fansink]

In terms of volume, potential server or AI accelerator orders placed by Microsoft or Amazon at Intel Foundry likely won’t be significant.

Assume Microsoft needs Intel Foundry to manufacture one million units per year of a particular in-house designed server CPU, and assume a 12-inch wafer yields 300 units of that CPU:


1,000,000 CPU units / 300 (units per wafer) = 3,333 wafers


3,333 wafers / 12 months = ~278 wafers per month


You can plug in your own assumptions, but you'll probably reach the same conclusion: the volume is not significant.

Translation...

18a is in serious trouble, everything after is suspect

Tan didn't wait til he heard the Death Knell , he’d’ve started long term fabless negotiations with TSMC, before he bacame CEO.