Info [WCCFTech] Intel CEO Wants To Destroy The Thinking About Having 90% Share In CPU Market, Talks 10nm Problems, 7nm Roadmap And More

Hitman928

Diamond Member
Apr 15, 2012
5,177
7,628
136
Normally I don't like to post from WCCF but this is mainly quotes from the intel CEO at a technology conference and they appear to be the only one who got some form of a transcript out so here is the link:


"We think about having 30% share in a $230 [silicon] TAM that we think is going to grow to $300B [silicon] TAM over the next 4 years, and frankly, I'm trying to destroy the thinking about having 90% share inside our company because, I think it limits our thinking, I think we miss technology transitions. we miss opportunities because we're, in some ways pre-occupied with protecting 90, instead of seeing a much bigger market with much more innovation going on, both Inside our four walls, and outside our four walls, so we come to work in the morning with a 30% share, with every expectation over the next several years, that we will play a larger and larger role in our customers success, and that doesn't just an CPUs.

Interesting that Swan is trying to change Intel's internal culture in regards to market domination but I'm not sure his viewpoint is clear? I mean the 90% remark seems only appropriate for x86 or maybe desktop/server/laptop at this point in time so I guess he's saying that he's trying to refocus the company on the most lucrative markets they can be the leader in rather than try to dominate every market? Or it could read that he wants to change the mindset of dominating the traditional x86 type market and focus more on new markets (e.g. AI, GPUs) where they can expand their TAM and increase their overall market presence/leadership.

Secondly, we're not going to try to do 2.4 scaling or 2.7 scaling as we think about 7 nm, you know, we put 2.0 back in line with historical trends as we think about 5nm, which would be our competitors 3 nanometer... Our first 7nm product [arrives] in the fourth quarter of 2021

Not much there except re-stating what I think we already knew. Some sites I've seen have taken this as their first 7nm CPUs will ship in 4Q21 but according to this quote, he just says product which was already expected to be a GPU if I am remembering correctly.
 

coercitiv

Diamond Member
Jan 24, 2014
6,150
11,670
136
Interesting that Swan is trying to change Intel's internal culture in regards to market domination but I'm not sure his viewpoint is clear? I mean the 90% remark seems only appropriate for x86 or maybe desktop/server/laptop at this point in time so I guess he's saying that he's trying to refocus the company on the most lucrative markets they can be the leader in rather than try to dominate every market? Or it could read that he wants to change the mindset of dominating the traditional x86 type market and focus more on new markets (e.g. AI, GPUs) where they can expand their TAM and increase their overall market presence/leadership.
My (subjective) take on this is something I said before on the forums: Intel lost it's ability to compete while replacing it with the arguably more profitable ability to dominate. Their strategy in every market (old, emerging, doesn't matter) is to establish supremacy and drive everybody out in a few years. The best example we got lately was mobile, they were incredibly aggressive, pumped billions of dollars, and when they realized it was going to be a long term effort they just gave up entirely.

If Intel's CEO has the vision (and the backing) to remap this giant's synapses into being a nimble competitor in different markets then I'm genuinely excited for their future. This is what Intel really needs, not some saving grace Conroe or Keller moment.
 

moinmoin

Diamond Member
Jun 1, 2017
4,933
7,618
136
Eh...
Secondly, we took on a 100% market share for smartphone modem and we decided that we would build it in our fabs, so we took on even more demand.
I'm repeatedly told that TSMC which does a lot of whole SoC dies for Apple, Qualcomm, HiSilicon (Huawei), MediaTek etc. can't handle the amount of dies required for the x86 market. And here we are with Bob Swan claiming Intel doing smartphone modems is what causes the repeated 14nm shortages Intel faces...
 
  • Haha
Reactions: Thibsie and lobz

firewolfsm

Golden Member
Oct 16, 2005
1,848
29
91
TSMC is projecting mass production of 5nm in 2H2020, while Intel is here stating Q42021 for their (nearly equivalent) 7nm. TSMC's lead should only increase beyond one year with 3nm.
 

lobz

Platinum Member
Feb 10, 2017
2,057
2,856
136
Eh...

I'm repeatedly told that TSMC which does a lot of whole SoC dies for Apple, Qualcomm, HiSilicon (Huawei), MediaTek etc. can't handle the amount of dies required for the x86 market. And here we are with Bob Swan claiming Intel doing smartphone modems is what causes the repeated 14nm shortages Intel faces...
That's just hilarious :D
 

lobz

Platinum Member
Feb 10, 2017
2,057
2,856
136
TSMC is projecting mass production of 5nm in 2H2020, while Intel is here stating Q42021 for their (nearly equivalent) 7nm. TSMC's lead should only increase beyond one year with 3nm.
That's just bull excrement from their side. They will be very happy if they can fulfill the Supercomputer contract at the end of 2021 with the PV chips. Very happy, I mean. I'd be shocked if intel's 7nm went into mass production any time before Q2 2022 (and by that I mean real high volume production for the consumer market).
 

liahos1

Senior member
Aug 28, 2013
573
45
91
TSMC is projecting mass production of 5nm in 2H2020, while Intel is here stating Q42021 for their (nearly equivalent) 7nm. TSMC's lead should only increase beyond one year with 3nm.

but nobody will touch that capacity in size in 2020 other than Apple. I think that is a distinction people here seem to forget. Binning 2 skus with a fixed performance target is different than what intel needs to do to get to full HVM on node across its portfolio
 

Gideon

Golden Member
Nov 27, 2007
1,608
3,563
136
but nobody will touch that capacity in size in 2020 other than Apple. I think that is a distinction people here seem to forget. Binning 2 skus with a fixed performance target is different than what intel needs to do to get to full HVM on node across its portfolio
Yes, but it doesn't matter, Intel won't have anything in comparable volume even in the last quarter of 2021.

Just read this article for comparison.

TSMC's N5 risk-production started in March 2019. N5P risk production starts in 2020Q2 (7% perf or 15% power improvement at iso-power compared to their 5nm). It will be their forth node using EUV. Real products using it will be out in 2020Q4 - 2021Q1 or about the same time that Intel get's limited quantities of their HPC GPUs out on their 7nm.

If all goes well, Intel's 7nm will be really ramping up in the middle of 2022. By that time, TSMCs GAA-FET 3nm will have been in risk-production for a couple of quarters.
https://fuse.wikichip.org/news/2879/tsmc-5-nanometer-update/
TSMC is taking smaller and more reasonable steps than Intel but still iterating faster overall. Intel failed hard with their 2.8x density scaling on 10nm. They could have learned from that in 2015-2016 and instead of going full-steam ahead inserted a simpler band-aid "10nm fix" node with EUV on some critical layers and smaller density increase (say <=1.4x to get back to the historic norm of 4.0x scaling from 14nm).

Instead they still went for 2.0x over the 10nm node (that's 5.6x density improvement compared to 14nm). That in-turn means the need for extensive EUV usage with zero prior experience with it and a broken 10nm node for reference. And that will be on a node that must make extensive use of EUV, arrive ASAP and fix the production issues of 10nm all in one go. What if Intel succeeds on 2/3 but 7nm turns out to be as manufacturable in 2022 that 10nm was in 2019?

In comparison, TSMC went with smaller density increases of around 1.8x . Even more importantly, they inserted complex new technology (FinFET, EUV) on the same node without increasing density at all.

wikichip_tsmc_logic_node_q2_2019.png


The risk of doing both Fin-FETs and a shrink paid off for Intel with 22nm, le'ts see how well 7nm does with EUV.
 

Gideon

Golden Member
Nov 27, 2007
1,608
3,563
136
I don't think TSMC is planning on using GAA-FET on 3nm.
Really? As Samsung is doing it (even a quite impressive version of it) I thought they are.

Well then It's gonna be tough to squeeze any power/perf improvements out of that node without it. Might end up as a node similar to their 20nm that was largely skipped until it got its 16nm makeover as there was almost zero power/perf benefit from 20nm alone only density.

Gels well with their usual attitude of inserting difficult new technologies on existing process nodes.

Contrast this to Intel, which needs to:
  1. Get the 7nm node out "yesterday" under immense management pressure to succeed
  2. Make it way more manufacturable than the previous node, from the get go while using tons of EUV for the first time
  3. Still scale 2.0x from the broken node, density wize
 

Gideon

Golden Member
Nov 27, 2007
1,608
3,563
136
A relevant comment from Reddit:
Smartcom5 said:
TSMC: We're in risk-production on 5nm since months, we're likely deliver first 5nm-chips to market by March '20 and HVM on 5nm is scheduled for Q2 '20, 3nm still on track for production in '22.

Intel: We can't fab enough chips on 14nm, never mind our broken 10nm, but we just started ordering 7nm EUVL-tooling and equipment in August, while the fab-expansion for 7nm in Kiryat Gat, Israel we've postponed indefinitely. Though, that means no·thing! As everything is still on track™ and 7nm is also totally ready to be launched in '21, that's why we ramped our 14nm by 25% this year. In fact, we're so excited about the ramp-up on 7nm, that we just backporting 7nm designs to 10nm now, since it looks way more promising to be fabbed on our totally working 10nm! Oh, and we're just started fabbing age-old CPUs on 22nm again, since we're totally thrilled of the progress me make on our 10nm and 7nm! Nothing to see here, you can trust us!
 

A///

Diamond Member
Feb 24, 2017
4,352
3,154
136
... we are going to drop well below 90% share and me talking this crap gives some illusion of us having a degree of control over that.
Makes sense. It's meant to pacify the investors. OTOH, it sounds like Intel would rather distribute their resources and not have all or most of their eggs in one basket. Then again, it being Intel I'm as confused as most people about their words and intentions.
 

krumme

Diamond Member
Oct 9, 2009
5,952
1,585
136
Intel needs a new BoD and a new far stronger ceo. The top president and directors layer needs to go, and the HR especially needs to go.
There is something fundamental wrong inside that org. There is no leadership and direction. Unfortunately the excellent economic results and numbers hide it. Because of decades with no compettition, sans a few years.
Its going to explode.
Look at that last fairytale ppt of a roadmap. What is that nonsense? Serious?
 

lobz

Platinum Member
Feb 10, 2017
2,057
2,856
136
Intel needs a new BoD and a new far stronger ceo. The top president and directors layer needs to go, and the HR especially needs to go.
There is something fundamental wrong inside that org. There is no leadership and direction. Unfortunately the excellent economic results and numbers hide it. Because of decades with no compettition, sans a few years.
Its going to explode.
Look at that last fairytale ppt of a roadmap. What is that nonsense? Serious?
Yeah I'm surprised noone opened a thread with the new 'node shrink every 2 years' process roadmap. https://wccftech.com/intel-2021-2029-process-roadmap-10nm-7nm-5nm-3nm-2nm-1nm-back-porting/
 

Ajay

Lifer
Jan 8, 2001
15,332
7,792
136
Really? As Samsung is doing it (even a quite impressive version of it) I thought they are.

Well then It's gonna be tough to squeeze any power/perf improvements out of that node without it. Might end up as a node similar to their 20nm that was largely skipped until it got its 16nm makeover as there was almost zero power/perf benefit from 20nm alone only density.

Gels well with their usual attitude of inserting difficult new technologies on existing process nodes.

Contrast this to Intel, which needs to:
  1. Get the 7nm node out "yesterday" under immense management pressure to succeed
  2. Make it way more manufacturable than the previous node, from the get go while using tons of EUV for the first time
  3. Still scale 2.0x from the broken node, density wize

TSMC has been working the node shrink problem with FinFETs with improvements to the transistor material/physical design. That how they got to 5nm with FinFETs. I really do wonder what they have up their sleeves to get FFs working on their 3nm node. They cannot afford to screw up as in the case with 20nm, IMHO - given the large number of high profile customers they have now.

Intel won’t be in high volume (HVM) on 7nm EUV till 2022. They have had a lot of time to refine their 7nm process, given the long delay in getting it out - so it's a good bet that it will have a much better rollout than 10nm (No stupid SAQP). Intel hasn’t said anything about 7nm's density or electrostatic characteristics AFAIK.

Further bad news for Intel is that TSMC will be in HVM on their 3nm in 2023 (last I saw). So AMD will still have a leg up on them, at least density wise.
 

CHADBOGA

Platinum Member
Mar 31, 2009
2,135
832
136
Intel needs a new BoD and a new far stronger ceo. The top president and directors layer needs to go, and the HR especially needs to go.
There is something fundamental wrong inside that org. There is no leadership and direction. Unfortunately the excellent economic results and numbers hide it. Because of decades with no compettition, sans a few years.
Its going to explode.
Look at that last fairytale ppt of a roadmap. What is that nonsense? Serious?
This is so on the money, it ain't funny.
6IU7awz.gif
 

Mopetar

Diamond Member
Jan 31, 2011
7,797
5,899
136
... we are going to drop well below 90% share and me talking this crap gives some illusion of us having a degree of control over that.

That's certainly true, but there was a similar speech that Steve Jobs gave when coming back to Apple where he said trying to compete with Microsoft was pointless. Part of it was a realization that there was nothing Apple could do about the situation, but the other part of it was to divorce the company's thinking from the notion that in order for Apple to succeed it meant that Microsoft had to fail. The company wound up being incredibly successful by finding their own niches instead of butting heads with the inevitable.

Granted, these points were made in different circumstances. Intel is the dominant player in this situation whereas Apple was a shadow of an also ran at that point, but I don't think anyone can deny that Intel has had a pig-headed attitude that has lead to a lot of failures over the years. Hubris is one of the most destructive forces in the universe and when you're such a dominant company like Intel has been it's often the case that you become a bigger obstacle to yourself than anything else as a result of all that misplaced pride.
 

scannall

Golden Member
Jan 1, 2012
1,944
1,638
136
That's certainly true, but there was a similar speech that Steve Jobs gave when coming back to Apple where he said trying to compete with Microsoft was pointless. Part of it was a realization that there was nothing Apple could do about the situation, but the other part of it was to divorce the company's thinking from the notion that in order for Apple to succeed it meant that Microsoft had to fail. The company wound up being incredibly successful by finding their own niches instead of butting heads with the inevitable.

Granted, these points were made in different circumstances. Intel is the dominant player in this situation whereas Apple was a shadow of an also ran at that point, but I don't think anyone can deny that Intel has had a pig-headed attitude that has lead to a lot of failures over the years. Hubris is one of the most destructive forces in the universe and when you're such a dominant company like Intel has been it's often the case that you become a bigger obstacle to yourself than anything else as a result of all that misplaced pride.
This may apply to Intel now. Quite prescient. Steve Jobs on Xerox
 

positivedoppler

Golden Member
Apr 30, 2012
1,103
171
106
Since Xerox FIRED me 17 years ago, I would love to know what was said in this video. I am deaf, and there is no cc options.

Please give me the quick synopsis

1. A successful company achieves monopoly
2. Upon achieving monopoly, marketing people are able to bring in more money vs product people.
3. marketing people with no sense of product or customer are promoted and gets to make even more decisions
4...I guess Xerox rotts from there
 

dmens

Platinum Member
Mar 18, 2005
2,271
917
136
Intel needs a new BoD and a new far stronger ceo. The top president and directors layer needs to go, and the HR especially needs to go.
There is something fundamental wrong inside that org. There is no leadership and direction. Unfortunately the excellent economic results and numbers hide it. Because of decades with no compettition, sans a few years.
Its going to explode.
Look at that last fairytale ppt of a roadmap. What is that nonsense? Serious?

Haha, it is already over. Just ask any of the torrent of smart and capable engineers who have left since 2016. As far as I can tell most people still at Intel are either coasting to the end of their careers, stuck due to family or community, or miking the cow until the bitter end because they are not employable elsewhere.
 

Mopetar

Diamond Member
Jan 31, 2011
7,797
5,899
136
Since Xerox FIRED me 17 years ago, I would love to know what was said in this video. I am deaf, and there is no cc options.

Please give me the quick synopsis

This isn't completely accurate, but it's a pretty good representation of the conversation. It seems like the conversation before this is in reference to Jobs visiting Xerox PARC for inspiration for the Mac.

Interviewer: "So they were very skillful?"

Jobs: "Yeah, but they did show us. So, and it's good that they showed us because the technology crashed and burned at Xerox."

Interviewer: "Why?"

Jobs: "I actually thought a lot about that and I learned more about that with John Scully later on and I think I understand it now pretty well. What happens is, with John Scully is that John comes from Pepsi Co. and they at most would change their product at most every ten years. To them a new product was like a new sized bottle. So if you were a product person you couldn't change that company very much. So who influenced the success of Pepsi Co.? The sales and marketing people. Therefore they were the ones that got promoted and therefore the were the ones that ran the company. For Pepsi Co. that was okay, but it turns out the same things can happen to technology companies that get monopolies. Like IBM and Xerox. If you were a product person at IBM at Xerox, so you make a better copier or a better computer, so what? When you have a monopoly market share the company isn't any more successful. So the people who can make the company more successful are the sales marketing people and they end up running the company and the product people get run out of decision making forums. And companies forget what it means to make great products. The product sensibilities and the product genius the brought them to that monopoly position gets rotted out by people running these companies who have no conception of a good product vs. a bad product. The have no conception of the craftsmanship required to take a good idea and turn it into a good product and they really have no feeling in their hearts about wanting to help the customers. So that's what happened at Xerox PARC. The used to call the people who ran Xerox toner heads. Toner heads would come out to Xerox PARC and they had no clue about what they were seeing."

Interviewer: "And for our audience what is toner?"

Jobs: "Toner is what you put into copier. The toner that you add to an industrial copier."

Interviewer: "The black stuff?"

Jobs: "The black stuff. They were basically copier heads that had no clue about a computer or what it could do. So they just grabbed defeat from the greatest victory in the computer industry. Xerox could have owned the computer industry. Could have been a company ten times its size. Could have been IBM. Could have been the IBM of the 90's. Could have been Microsoft of the 90's."
 
  • Like
Reactions: ehume

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,478
14,432
136
So... I worked for the printer division at Tektronix. They made solid ink printers. Amazing photo printouts on plain paper ! And if you used photo paper, its was mind boggling. So what Xerox did was buy out that division, since it was hurting their sales. But instead of continuing those printers, they killed off solid ink. I think that is one of the things that may have killed them.
 
  • Like
Reactions: A/// and ehume