TSMC’s 3nm & 5nm Processes Expected To Generate Over $31 Billion in First 3 Quarters

[jorgequinonez]

Who knows. It may turn out to be a blessing in disguise if Intel is forced to spin off its manufacturing division. AMD and Globalfoundries' split created a very successful AMD and a not-too-bad Globalfoundries. Most important thing is that both of them survived.

Intel does face a crisis but it can use it as a turning point for changing to a better and viable business model.

Don't you think that spinning off the manufacturing division would appear more like a negative move, as it could be perceived as a sign of failure by the market?

 

[hist78]

Don't you think that spinning off the manufacturing division would appear more like a negative move, as it could be perceived as a sign of failure by the market?

No, I don't think so. After the split, there is a great chance that one of two or even both of them will succeed. Without changing Intel's business model, both the Intel design division and manufacturing division can be wiped out by the competitors. Right now Intel should put all its attention to its survival instead of its stock price. Actually Intel put wrong amount of attention in keeping its share price high is one of the reasons that cased Intel's troubles today.

 

[hist78]

No, I don't think so. After the split, there is a great chance that one of two or even both of them will succeed. Without changing Intel's business model, both the Intel design division and manufacturing division can be wiped out by the competitors. Right now Intel should put all its attention to its survival instead of its stock price. Actually Intel put wrong amount of attention in keeping its share price high is one of the reasons that cased Intel's troubles today.

By looking at Intel's share buyback data, I feel for the past 20+ years Intel's CEOs, CFOs, and Boards of Directors did terrible jobs without vision and honesty.

Since year 2000, Intel spent $133.859 billion on buying back Intel shares in order to push EPS and share price higher.

OR

Since year 2010, Intel spent $83.535 billion on buying back Intel shares.

OR

Since year 2018, Intel spent $40.947 billion on buying back Intel shares.

After Q1 2021, Intel ran out of money and stopped the share buyback program. What did those huge amount of cash outflow achieve? How many missed opportunities, wrong strategies, and poor decisions were made due to this lavish illusion?

Now Intel is struggling to keep its operating cash flow positive! How sad it is.

 

[osnium]

By looking at Intel's share buyback data, I feel for the past 20+ years Intel's CEOs, CFOs, and Boards of Directors did terrible jobs without vision and honesty.

Since year 2000, Intel spent $133.859 billion on buying back Intel shares in order to push EPS and share price higher.

OR

Since year 2010, Intel spent $83.535 billion on buying back Intel shares.

OR

Since year 2018, Intel spent $40.947 billion on buying back Intel shares.

After Q1 2021, Intel ran out of money and stopped the share buyback program. What did those huge amount of cash outflow achieve? How many missed opportunities, wrong strategies, and poor decisions were made due to this lavish illusion?

Now Intel is struggling to keep its operating cash flow positive! How sad it is.

I can understand why taxpayers are angry at having to give Intel $8.5B given the misuse of money.

 

[IanD]

The CFO has mentioned this before 18A is not meaningful until 2026 or later.

What is the volume you could run in D1x?

Which is exactly why I said the real competition for volume CPU production -- what really matters -- with BSPD is 18A for Intel vs. A16 for AMD in 2026...

Small-volume sampling earlier than this makes headlines and a "product launch", but this is exactly what Intel did in 10nm and then failed to ramp up.

If they can't ship in volume with good yield and competitive costs then AMD/TSMC will win the "process" battle, even if Intel technology is as good (or even better!) for PPA.

 

[siliconbruh999]

I think you are misinterpretating TSMC/AMD timeline there ain't no way AMD gets first access to TSMC before Apple TSMC Says 2H26 for A16 meaning at least 2H27 for AMD to get anything for 2nm it is 25 so 2H 26 for AMD regarding yield only time knows the answer not even Intel/TSMC/AMD also for 18A we have to see if Intel 3 is anything to go by it is the only node in the last decade to be on time from Intel

 

[IanD]

I think you are misinterpretating TSMC/AMD timeline there ain't no way AMD gets first access to TSMC before Apple TSMC Says 2H26 for A16 meaning at least 2H27 for AMD to get anything for 2nm it is 25 so 2H 26 for AMD regarding yield only time knows the answer not even Intel/TSMC/AMD also for 18A we have to see if Intel 3 is anything to go by it is the only node in the last decade to be on time from Intel

Did you actually read what you posted?

"A16 is best suited for HPC products with complex signal routes and dense power delivery networks, as these can benefit the most from backside power delivery".

What they don't say is that there's also quite a big cost premium for A16 -- and yes, I've actually had discussions with TSMC about all this instead of just reading press releases or what's posted on the web.

For all those reasons it's unlikely that A16 is a good fit for Apple mobile CPUs, it's much more likely to be used by Nvidia and AMD (when -- 2H26?) -- and maybe even Intel, if Intel Foundry can't deliver 18A as promised... ;-)

(where "delivered" means "in high volumes with good yield and competitive cost", not the meaning Intel (mis)used for 10nm)

And that's the kicker for Intel, not any theoretical dates about "process availability". If I was a betting man I'd put my money on A16 having more customers in much bigger volume sooner than Intel CPUs in Intel 18A -- and that's the measure of success for a foundry, not "I got there first!"...

Don't forget that A16 is pretty much the transistors from N2/N2P (which Apple certainly *will* use) with BSPD added on for HPC applications -- and as a consequence a completely new set of IP is needed, which makes it less attractive for anyone who has already done N2 developments (e.g. Apple). My guess is that Apple will use N2/N2P and then N2X (all topside power), and then switch to BSPD for A14 by which time the cost penalty will also have come down since A16 will have pipecleaned the technology.

BSPD is in a very similar situation to when TSMC introduced EUV initially in N7+ before going all-in on N5 -- the technology change meant N7+ layouts were incompatible with N7 (and N7+ cost more) so new IP was needed and it got very few users, but it enabled TSMC to get the kinks out of EUV and then bring the cost down and volume up for N5.

 

[DepthFirst]

That’s an easy bet. It would require some sort of catastrophic failure for TSMC to not ship a larger volume than Intel at any point in this decade. Even if Intel was quite successful with their fab ambitions.

 

[IanD]

That’s an easy bet. It would require some sort of catastrophic failure for TSMC to not ship a larger volume than Intel at any point in this decade. Even if Intel was quite successful with their fab ambitions.

Maybe not at first for 18A if Intel get there before A16 and can ramp it up fast. But even if they manage this (which I doubt, but let's see...) I suspect A16 will soon overtake it because -- well, Nvidia... ;-)