Hacker News Reader

Comment by lizknope | Hacker News Reader

Intel wants to get into the foundry business. 99% of what Intel does is for their own chips. They don't have the support infrastructure of being able to deal with external customers. Global Foundries has that.

You don't just decide to make a chip. There are hundreds of foundry specific libraries, design rules, and more. There are always problems with the EDA CAD tools for new processes and you need lots of support people to answer questions and help debug problems that your customers are having.

Intel has tried being a foundry before multiple times over the last 20 years. They always fail and give up after a couple of years.

View on HN

Well, Globalfoundries was exactly in the same situation directly after they were spun off from AMD and they bought Chartered Semiconductors to get foundry knowledge into their company. They also hired a lot of ex-freescale and ex-IBM people.

Wait I thought GloFo essentially gave up the node race and was settling into 'not the latest and greatest' market a la TI. Is that not a correct assumption?

Essentially, but now they have Intel nodes to sell as well. It's a fairly reasonable plan, and I doubt Intel will stay behind TSMC node-wise forever.

If Intel ends up buying AMD's (former) business, then damn the tables have turned.

Considering the way things are going I am more than a bit intrigued, but not surprised.

I am waiting for the headline "AMD to buy Intel!"

And Nvidia buys ARM, which is then bought by AMD-Intel, which then merges with Samsung.

And then The Semiconductor Manufacturing Company takes shape.

and then Apple buys Samsung

and then Amazon buys Apple.

and then Aliexpress buys Amazon.

And Softbank already owns Aliexpress. And then CCP intervenes!

And then the CCP breaks it up into the original pieces.

The things I've seen, I would be less than surprised this preposterous plot plays out.

Given Intel wants to sell fab space it wouldn’t shock me if AMD keeps buying fab space at a soon to be Intel plant to make chipsets and other items that don’t need to be on the bleeding edge or constitute sharing much IP.

That would be a terrible outcome competition-wise :(

>>before multiple times over the last 20 years. They always fail and give up after a couple of years.

That would be the concern. What happens to the chip market when intel rediscovers its inherent anti-competitive nature and shuts down any non-intel product?

Intel is already in the foundry Business their operating nodes are very large 14nm and 10nm (Which is 12nm) and 7nm process node is broken due to poor yields.

And the tech/IP they offer is way behind the competition. They were not a realistic choice when we last sat down.

But the chance to utilise EMIB would be cool.

Reply on news.ycombinator.com