Thread regarding Intel Corp. layoffs

IDM and IFS is a failed strategy.

This isn’t the 90s or 2010 when silicon RD and factory CapEx was in the billions and tens of billions and when x86 volume and complexity for the process was simple enough to enabled Intel to be a leader and to have GM the envy of the industry.

Times have changed about a decade ago and that was missed by the BoD and executives, past the point of recovery.

Today leading edge nodes require tens of billion in basic and ecosystem development for a node and easily 30-50 billion for the fabs. X86 can’t pay for it any more as it doesn’t have captive revenue nor margins nor volumes to sustain that expenditures.

So snake oil Pat say we combine that with Foundry. And have the combo and we conquer the world and get back to leadership.

So many reasons this will fail 0). Technology and physics is the same in Hillsboro, Hsinchu and Seoul. Intel can’t get ahead unless the others fail as spectacularly as LTD did at 10nm. 1) service minded DNA is non existent at Intel and IFS will fail on that alone 2) infrastructure to support the varied product mix and volume for IFS nonexistent. TMG has no clue how to do this nor the skill set. 3) lagging technology so products will be uncompetitive and priced low and 4) lack of scale which combined with few products results in slow leaning, low yields and inefficient cost structure. Lastly the elephant in the roomNO customers everyone that use leading edge node competes with Intel. Except maybe DoD which is low volume as it gets. Who goes to your direct competitor Fabs for chips and they see your Yield, layout and marginalities and also make their own direct competing part. You think if they see a weakness they tell you or their internal guys?

No IFS and IDM is fundamentally conflicting to the fabs, and Intel has no scale or talent to do this separate. Perhaps 15 years ago under Bills watch when we tried it and really did it could have happened, but that window has come and gone and closed.

Now to try and do both will result in a worse outcome than if they even just tried just one. Just IDM does’t make sense for the reasons I noted, behind, lack of scale means never competitive, and just IFS the same.

Now you try to do both when you already failed so badly at 10nm sounds like fantasy from the leadership.

Pat and the executives up to fellow and VP don’t care l. They work, tell a good story and get paid so much if it goes Kodak belly up they retire. It’s tax payer and stockholder that get screwed no big deal to them, so they’re payday a lot smaller but they can retire still!

How about you poor < Gr 10 and < 50 you got a plan ?

by
| 1719 views | | 1 reply (last ) | Reply
Post ID: @OP+1kqvxX9E

1 reply

Meteor Lake is on Intel4 and Intel’s claim to relevance again, EUV and competitive to the Foundrys. Pat fails to mention TSMC has been running 5nm for two + years and 4nm for over a year at huge volumes, millions of wafers l and has an insurmountable lead in process and tool learning. No amount of tool manufacturing help makes up for running real wafers.

AMD, Nvidia and others all get this superior advantage over an Intel internal IDM, crazy for anyone to spin that the IDM has an advantage!
Maybe there was 8 years ago on 14nm that was a rational argument, now it is snake oil salesman when you are behind and have no scale.

by
| | Reply
Post ID: @1ugq+1kqvxX9E

Post a reply

: