News Morris Chang Asserts Intel Foundry Will Remain in TSMC's Shadow

Status
Not open for further replies.
Intel has "Copy exactly." and has been very successful with it.

TSMC should follow the same plan they have been. "Copy themselves exactly". By that I mean keep doing exactly what they have been doing and "run their own race" without looking back.

I still think that microprocessor growth is going to continue to be so explosive that TSMC and Intel can both grow enormous amounts and may both have trouble building all the capacity they can use.
 
Foundry business will get very tough because they are growing at a rapid pace, yet at the same time, their client base is shrinking due to geopolitics tension resulting in sanctions.
For Intel to catch up, there are 2 things that they need to address,
1. Talents - I think this is where TSMC and Samsung have an edge. You can't just get someone on the streets to do these stuff even if you have new fabs.
2. Cost - With cost rising, I feel Intel being predominantly in the West is not able to compete effectively when it comes to pricing. So in my opinion, Intel is likely securing orders from companies that want to diversify just in case.
 
It's impossible for Intel to ever compete with TSMC in anything but leading edge nodes. TSMC still has a lot of 200mm wafer capacity while Intel doesn't, and building said fabs are not viable which is a large part of the reason Intel was trying to buy Tower.

We've seen a lot of really high profile diversification when it comes to leading edge nodes. They're sticking around a lot longer than before and companies aren't always hopping on the latest and greatest across the board. This is where Intel has a lot of room to compete as they just have to be able to provide capacity at a reasonable cost.
 
I think I interrupted the translation different than what is in English.

"... it does not pose a substantial threat to TSMC unless it manages to enhance various facets of its foundry operations like technological leadership, yield rates, and competitive pricing. Meanwhile, even if Intel Foundry Services is successful, it will still be a shadow of TSMC, according to Chang."

My interruption is

" when looking as the competition from Intel, Chang is not worry, he point out, if intel have good service, technology is catch up, yield is improved, with the same price, then this is only a shadow of impact to TSMC"

If a competitor can provide a better service/product to (i.e. Apple) customer, with a lower prices, Then customer will not to jump boat?

This is what Intel has in the moment, their Intel 4/3 is just as completive as TSMC N3E, and Intel 20A/18A will come out of TSMC more then 2 years of their equivalent product i.e. TSMC N2P, will the customer jump ship, it is more likely then not.
 
Until intel spins off foundry services I don’t see it catching up with TSMC in terms of generating customer interest. Why would Qualcomm or AMD or even Apple trust intel to build chips for them, when intel is directly competing against them in laptop, desktop, server, graphics? There’s a massive underlying conflict of interest there. TSMC on the other hand is a neutral party.

That being said, there needs to be a competent alternative to TSMC. TSMC simply doesn’t have enough capacity online to satiate market demand especially on the bleeding edge nodes.

So it’ll be interesting to see whether intel’s massive $$$ spending on foundry capacity will pay off in the market place, or whether much of it will sit idle, being a waste of billions of dollars. Perhaps is shrewd enough not to spend billions without there being any market/customer demand for its services.
 
Until intel spins off foundry services I don’t see it catching up with TSMC in terms of generating customer interest. Why would Qualcomm or AMD or even Apple trust intel to build chips for them, when intel is directly competing against them in laptop, desktop, server, graphics? There’s a massive underlying conflict of interest there. TSMC on the other hand is a neutral party.
I see so many people spouting this nonsense that I can't tell if it's irrational fear of Intel or just complete ignorance as to how business and chipmaking work. The turnaround time for architectures is so long that even if they were dumb enough to steal something it'd be years before they reaped the benefits. Then there's the fact that if they did it in the first place their entire business would be blackballed and effectively dead. These are just the two most obvious reasons as to why this is a dumb thing to say.
 
  • Like
Reactions: RandomWan
Status
Not open for further replies.