[alecco]

> That's what I'm trying to understand. His educational background was in Physics/Nuclear Engineering so he's obviously a smart guy, and he was CEO/Chairman of Cadence for 15 years, but other than that his 40+ year career has most been in VC and being on the boards of an incredibly large number of companies.

He is no Pat. He is no Andy. He is a business guy with some hard science behind (not electronics per se). It doesn't feel right.

[ksec]

>He is a business guy with some hard science behind (not electronics per se). It doesn't feel right.

I think you need to look up Cadence and look into how the fabless industry works. Picking him means Intel is possibly about to spin off or spin out the Chip division and only focus on Fabless.

[ahartmetz]

My interpretation is that Intel didn't understand the customers of the fabless business and Tan does, so he's there to make Intel fabs attractive to them.

[bgnn]

EDA companies have limited exposure to the inner workings of the fabless semi giants (Broadcom, Qualcomm, AMD). They have even less exposure to running a fab service for fabless companies.

[bcrl]

EDA tool vendors have to work extremely closely with both fabless firms and fabs forever to tune their products to deal the ever increasing complexity of nanometer scale manufacturing processes. Backside power delivery? You bet your ass that the tools folks were involved in making that work well for designers. Gate-All-Around? Probably needed the tool vendors to make tweaks based on feedback from the fabs.

The important thing is that Lip-Bu is from the industry, and has contacts on the tools side of things as well as the customers for those tools which happen to be potential future customers for any Intel fab services. This is a step in the right direction for INTC which has a board where industry experience is severely lacking.

[ahartmetz]

Yes. There is an ever increasing amount of complicated design rules for modern processes and EDA tools have to model all of them for automated layout.

[nickpsecurity]

Anyone that's successfully been running hardware companies would be where I'd start. Offer the best of them much more money than they're currently making. Actually, hire a dream team of them. Turning Intel around would be worth paying a high price.

[tacticalturtle]

Lip-Bu Tan would fit that criteria.

Cadence is a successful hardware company - they have an IP catalog available for licensing and sell design services.

[nickpsecurity]

Maybe. It's a good start. The part where I'm unsure is that Cadence is one of the Big Three vendors. They have so much I.P., EDA tooling, patents, etc. that money will keep rolling in with incremental innovation.

Most hardware startups have to come up with new ideas, engineer good implementations of them, market them, and react to all kinds of competition. I'm talking companies like Habana or Cavium more than Cadence.

When I say hardware CEO, I meant one that ran a company like that where we know they'll be innovative and capable. Intel will need that since they're going to have to change so dramatically with so many product innovations. Whereas, running Cadence doesn't require (from what I know of them) that sort of innovation and it might have even hurt their entrenched position.

That said, I'm still positive about the new CEO. I hope his mindset and experience helps him do a lot of good. The letter he put out was great.

[markus_zhang]

Given the current situation of Intel, it might need a lot of house cleaning to make any strategy work.

[gkanai]

> It doesn't feel right.

Who would you have as CEO?

[UncleOxidant]

Gelsinger. Should've kept Pat and gotten rid of (at least some of) the board.

[brewdad]

His problem was that he had a vision, possibly the correct one, with absolutely no clue how to execute it. That is unforgivable in a CEO.

[UncleOxidant]

Sure he knew how to get there: build SOTA fabs that can compete with TSMC. The problem was that it was going to cost $100B and take 10 years. The board wasn't patient enough and wanted a quick fix. There is no quick fix.

[YetAnotherNick]

They already spent 10 years and probably $100b. 14nm with Broadwell was launched in 2014 and they never recovered from there. They just launched one node process in 2019 in full capacity after that and by that time it was too outdated.

There is no way to justify that a market leader would fall this bad this fast and couldn't recover and loose billions in process, except something was terribly wrong.

[fc417fc802]

> fall this bad this fast

That's just how fabs are. Yeah, something went terribly wrong. That much is obvious.

Regardless of who was picked as CEO getting back to the cutting edge was always going to be horrendously expensive and take on the order of a decade. There was never a magical unicorn CEO that could avoid that.

The only obvious (at least to me) alternative is to sell off the foundry business. But someone is going to be running a cutting edge process at scale in the US. The federal government will presumably see to that.