Spoiler alert -BIG FAT NOPE!
LOL ICF is a joke.
5 replies (most recent on top)
ICF is a joke, another ICG (remember Maloney, Christensen........ anyone?). They do not know how to achieve goals the customers want. One size fits all is their matra which has already doomed them. They just don't know they are already doomed. Sooner ICF closed the better for Intel stock holders
Every internal team that has used TSMC still prefers TSMC over the totally customer service oriented TMG team.
Just go ask anyone that has designed their what it is like to have to come inside. Dang even the inside teams would love to go outside.
With Murthy in-charge I can see a push, it would result in faster design to production and be a wake-up call for TMG.
The only way they can get customers is to give away the process
Intel knows it can't compete on that front now anyway.
I think it already gave up.
Intel ICF lost customers after realizing how shi**y is their way if doing things. This is the result of a crooked culture.
This foundry thing is a chicken or egg thing. LIttle folks with a few wafers starts can be supported.
Qualcomm or the fruity company that need thousands upon thousand of wafers on timely predictable cadence.. NOPE!
1) Capacity needs to be put in place years ahead. Intel already built the factories, of course there is that empty shell in Arizona, LOL.
2) Tools like litho and some etchers have two year lead times. Sorry no additional slots for Intel. TSMC and Samsung are bigger customers these days. Go see how BIG their fabs are compared to intel.
3) Intel will NOT displace its own x86 products that sell for 100s to 1000s of dollars / die for Qualcomm or Fruity company that wants to only pay 20-50 bucks for their silicon.
Thus Intel is in a quandry.. it needs to commit another 4-5 billion in capital up front for 10nm NOW that won't be ready till 2019 time frame for any significant volume on 10nm. By that time anybody can go to TSMC and Samsung and get 10nm for a few thousand dollars / wafer cost on a process that is easier to design in .. BK is too little too late.
A few years ago there was argument that an IDM had an advantage working closely with design to make best silicon/product co-optimization. These days ARM is working with TSMC and Samsung on 10nm, 7nm, and soon 5nm developing library and IPs as are all the IP manufactures. TSMC and Samsung have done this now for multiple generations and now how to say yes to their customers. Intel unless the Zebra changes it's stripes is still arrogant and finds it hard to say yes to it's customers.. So customers may tease intel in the beginning but will quickly go elsewhere, I know that for a fact, LOL