|
I'd say they're better off than AMD was back than because the *cove cores are looking to be at least decent even they end up held back by the process + Intel doesn't have to rely on a 3rd party foundry to make them. As near as anyone can tell it looks like AMD will be highly fab limited forever at this point so even if they had a chip that out performed Intel's by 1000% Intel would be able to still stay in business for quite a while since AMD can only supply a fraction of the x86 market.
|
#
?
Jul 28, 2020 06:08
|
|
|
|
| # ? Apr 19, 2024 02:41 |
|
|
SourKraut posted:You got proven wrong on your arguments about the 10 nm/ 7 nm timelines, and instead of just acknowledging it, you keep trying to argue away that it wasn't a delay because Intel updated their roadmaps/etc. But no one except you is associating it with impacts on the stock market, investors, etc. And if Intel initially indicates that 10 nm is out in 2016 and 7 nm would be out in 2017, as was shown, and then update it between initial presentation and those dates to show it's now 2018, 2019, 2020, whatever, it doesn't matter what the reason for the delay is, because it is still a delay. Because we are talking about whether or not Murthy is getting scapegoated for the delay and if Bob deserves more blame. I'm not arguing over the definition of what a delay is. I'm arguing that this is the delay that got investors pissed and the one Bob took swift action on. The previous delays didn't cause a huge amount of market cap to disappear over night and the explanations for the delays either didn't come as a surprise or weren't as alarming as this one. Investors can deal with delays and roadmap changes as long as they are communicated far enough ahead. Drop a surprise bombshell and they get pissed. Again, I think this was recent news to Bob and he communicated this fairly promptly to investors and took prompt action afterwards in response to investor outrage. I'm not really expecting him to do more than that right now. If we find out that Bob knew all of this a year ago and was stalling or whatever then I'll apologize quote:Ultimately it doesn't matter what you believe, because as you continue to like to point out, he is a "finance guy", and so he's going to strive to maximize shareholder value, just as BK did... until it cost him. Which isn't to say that BS will do the same, and hopefully he demands and provides full transparency going forward. I point out he's a finance guy because he can't fix this and as much as people say he should hold people accountable  if you think a bunch of career engineers can't figure out a way to fudge or paint a pretty picture. Bob doesn't have the knowledge to call people out on this until it's become completely obvious they are lying and that's where we are at now. He can and should purge but he doesn't know enough to do that unless he burns the whole thing down, which he was probably trying to not do to avoid investor panic in the first place. And he doesn't have technical experts to help him either because Murthy was supposed to be that guy, Keller bailed, and everyone else left is career company men who you don't trust anyway. quote:But ultimately, you seem to have some type of emotional investment in defending Bob Swan/Intel I've repeatedly said Bob is not the guy who can fix this but I guess saying "the board hired the wrong type of guy to fix this" is considered emotional investment. If the board hired him to doggedly keep the fabs on track and fix whatever was wrong with it, then they were absolute fools and hired the wrong guy. quote:That was me, and I feel it was justified and even said as much by saying he deserved it to some extent, this is a problem that goes far deeper than Murthy alone. Maybe Bob Swan can start by changing Intel's toxic work culture that has driven everyone I know who used to work on their fabrication engineering teams away over the last 8 years. Bob is a numbers guy and I don't think a numbers guy is going to go in and fix the #1 complaint most Intel employees have. WhyteRyce fucked around with this message at 07:02 on Jul 28, 2020 |
|
#
?
Jul 28, 2020 06:26
|
|
|
Krzanich came from the manufacturing side and was a terrible ceo (except to investors) and set the company up for failure. I'd rather have a cfo in the role who can pick good leaders than a poo poo engineer ceo
|
|
#
?
Jul 28, 2020 06:32
|
|
|
Malcolm XML posted:Krzanich came from the manufacturing side and was a terrible ceo (except to investors) and set the company up for failure. BK did some great stuff like Intel Sports or that TBS makers show. One of his huge failings was that he fell into thinking he was one of the tech rockstar CEOs WhyteRyce fucked around with this message at 07:06 on Jul 28, 2020 |
|
#
?
Jul 28, 2020 06:41
|
|
|
PC LOAD LETTER posted:As near as anyone can tell it looks like AMD will be highly fab limited forever What does "fab limited" mean?
|
|
#
?
Jul 28, 2020 07:05
|
|
|
punk rebel ecks posted:What does "fab limited" mean? AMD doesn't manufacture their own chips: it has TSMC fab-ricate it for them This means that no matter how much better AMD's chips are than Intel's, Intel is always going to have some segment of the market available to them, just from people who can't be supplied by AMD because TSMC can't produce AMD's products fast enough
|
|
#
?
Jul 28, 2020 07:08
|
|
|
punk rebel ecks posted:What does "fab limited" mean? gradenko nailed it. I actually think AMD is more fab limited now than they were back in the K6 days. Back then they could supply something like ~25-30% of the x86 market. What they can do with Zen2 and up cores is either take big ol' chunks out of the x86 server market from Intel or force them to drop prices a bunch which would hurt quite a bit for Intel's financials but the company wouldn't be at risk of going under.
|
|
#
?
Jul 28, 2020 07:18
|
|
|
PC LOAD LETTER posted:gradenko nailed it. Do you think there's a possibility of Samsung being able to ramp up enough to handle consumer 7nm for AMD? They already announced that partnership together with a while back for the new Exynos SoC with Radeon and with TSMC now becoming the most crowded apartment building in the semi-conductor industry; I think it might make sense. https://www.reuters.com/article/us-samsung-elec-intel-stocks/samsung-electronics-shares-rally-on-intels-chip-outsourcing-plan-idUSKCN24T0C5 quote:Samsung Electronics is taking on TSMC in the chip contract manufacturing business, announcing last year a plan to invest 133 trillion won ($111.43 billion) in non-memory chips through 2030.
|
|
#
?
Jul 28, 2020 07:27
|
|
|
I don't know enough about Samsung's fab capacity to say for sure but I think AMD and pretty much everyone else looking to make parts on a post 14/12nm process else would almost have to be assesing them as a supplier if for no other reason than as a back up in case TSMC runs into issues.
|
|
#
?
Jul 28, 2020 07:31
|
|
|
Intel moves to TSMC 7nm to achieve parity and steal fab capacity double whammy for AMD
|
|
#
?
Jul 28, 2020 07:40
|
|
|
gradenko_2000 posted:AMD doesn't manufacture their own chips: it has TSMC fab-ricate it for them I see. Makes perfect sense. Thank you.
|
|
#
?
Jul 28, 2020 08:15
|
|
|
It’s pretty clear Intel needs to clean house. I don’t know enough to say whether these are the responsible guys but Intel can’t just let things keep slipping. They’ve slipped four years in four years on 10nm and now it’s happening again on 7nm. This is the one slip they can afford, and additional slips are all but assured if they massively clean house. They can maybe pull off one more 6 month slip if they stretch it, but they are already way behind AMD (which isn’t even using leading nodes at TSMC). Not even sure how you fix this mess since it’s basically a problem of production costs and it’s baked into the process they’ve chosen and the designs they’ve Implemented in those processes. Physics don’t care that someone got fired. I guess the simple answer is just ear the costs even and run the process even if the margins aren’t what you like, and continue doing R&D to try and bring yields up, but that depends on how bad things are. Comedy option, buy NVIDIA, fire Raja, make Huang the CEO of the merged company. They gotta find someone with engineer-CEO chops like Su or Huang, or at least someone like Keller who understands the engineering and tells it like it is. Maybe there’s some process guy they can acquire from TSMC or GF or Samsung or something. Then try to hire back as many of the people you drove away over the last 10 years as you can (good luck). Paul MaudDib fucked around with this message at 09:03 on Jul 28, 2020 |
|
#
?
Jul 28, 2020 08:59
|
|
|
Intel still makes something like 9 out of every 10 x86 CPU's made, to put into perspective how much of a manufacturing advantage they have over AMD. And Intel sells all those CPU's as fast as they make them, demand for CPU's has been at record levels. Even with inferior products, as long as they are 'good enough', which Intel CPU's still are, they'll still sell whatever they make as fast as they make. Intel falling behind on performance has forced them to cut prices significantly though, so people don't just take the option of waiting for AMD parts to become available. It's extremely fortunate for Intel that demand for new hardware is at record levels and looks to continue that way for the time being, it makes waiting in line for AMD products less appealing. Windows 7 going EOL has forced many corporate entities to refresh their hardware to get stuff that runs Windows 10 properly, now coronavirus is causing a surge in sales for newly deployed WFH equipment.
|
|
#
?
Jul 28, 2020 09:06
|
|
|
Yeah I was about to say that (I think) a lot who Intel sells to are business/office machines, for which 14nm (hell, 22nm) is probably going to be more than good enough. If you "just" need a laptop/desktop to push MS Office and Skype/Zoom/Teams, even a 2c/4t CPU can hack it, with lack of memory probably being a bigger issue for most use-cases.
|
|
#
?
Jul 28, 2020 09:33
|
|
|
gradenko_2000 posted:Yeah I was about to say that (I think) a lot who Intel sells to are business/office machines, for which 14nm (hell, 22nm) is probably going to be more than good enough. If you "just" need a laptop/desktop to push MS Office and Skype/Zoom/Teams, even a 2c/4t CPU can hack it, with lack of memory probably being a bigger issue for most use-cases. Very much this. The enthusiast PC market is a tiny, tiny fraction of the overall PC market--albeit a profitable fraction. In the same way that a ton of businesses are still ticking the box for a 1336p laptop screen because it saves them $15 vs the 1080p option for their employees, that an AMD processor is 10% faster at a given level means literally nothing to the people buying lots of 500 systems for a tech refresh if it's not also 10% cheaper. And that's why AMD wasn't stupid and took their Threadripper shot at HPC/datacenter first: that's where the real money is, and that's where people will actually care about relative performance metrics. So while Intel might be crying all the way to the bank at the moment, they do need to figure their poo poo out or they're going to start taking some major sales hits in the datacenter, and there's no way that doesn't show up in profit reports.
|
|
#
?
Jul 28, 2020 15:11
|
|
|
Paul MaudDib posted:It’s pretty clear Intel needs to clean house. I don’t know enough to say whether these are the responsible guys but Intel can’t just let things keep slipping. They’ve slipped four years in four years on 10nm and now it’s happening again on 7nm. They have to go about 2-3 tiers below where Murthy was to cut out enough of the rot.
|
|
#
?
Jul 28, 2020 15:45
|
|
|
wet_goods posted:They have to go about 2-3 tiers below where Murthy was to cut out enough of the rot. They've gone 2 layers before. Hell, murthy was an entirely new layer himself. There are plenty of department heads, tech leaders, and even first level managers in a company that large which can keep the rot persistent and toxic culture ingrained.
|
|
#
?
Jul 28, 2020 16:08
|
|
|
wet_goods posted:They have to go about 2-3 tiers below where Murthy was to cut out enough of the rot. While being careful not to cut out exceptions to the rule? Total headchopping isn't really as great of an idea as it seems on paper, especially if documentation and other efforts to make such a transition happen result in a brain drain of what good knowledge there is left. edit: if you go too far in removing a culture, you can end up with brain drain of administrative know-how at the very least (think: not just brain drain, but connections to SMEs on various subjects or at least access to those SMEs) and general conditions bad enough to instill a new culture that's just as bad Sidesaddle Cavalry fucked around with this message at 16:17 on Jul 28, 2020 |
|
#
?
Jul 28, 2020 16:12
|
|
|
Yeah. You are all right. Intel as a deathgrip on computers. I feel if any CPU rival will topple them in market share it will likely be an ARM manufacturer due to computers becoming absurdly small and/or integrated.
|
|
#
?
Jul 28, 2020 16:18
|
|
|
The specific form-factor of a desktop (as in a keyboard, mouse, monitor, on a desk) or a laptop is here to stay, but just the other day I was already idly musing about getting a Raspberry Pi 4 just for the novelty of using it as a forums posting machine
|
|
#
?
Jul 28, 2020 16:23
|
|
|
What I'm trying to say is you can go overboard with the populist draining the swamp poo poo. It doesn't matter how many hyperactive geniuses you throw on a team and management team - engineering enormous loving projects like general purpose microprocessors is a team sport, and if those geniuses don't have the skills to put their work together and/or keep their own poo poo straight, you'll get anything from a last-minute kludged-together design to a bunch of little useless products looking for a solution. And if you don't even leave enough people to pass on the torch, you kneecap the new team on top of that
|
|
#
?
Jul 28, 2020 16:29
|
|
|
punk rebel ecks posted:It's crazy how thing have managed to completely flip, but the time for it to be possible checks out. I can't believe it's been almost ten years since the Bulldozer fiasco. I wonder if Intel is in a worse position now than AMD was in 2011? I think I'd compare it more to Pentium 4 vs. Athlon 64. Intel is still able to (for now) brute force their way to the top of some benchmarks if you don't care about power and heat, and is still the bigger company sold in more systems. There's no excuse for computer enthusiasts to buy Intel but some will anyway.
|
|
#
?
Jul 28, 2020 16:34
|
|
|
Nomyth posted:What I'm trying to say is you can go overboard with the populist draining the swamp poo poo. A common dilemma anywhere is a keeper of tribal knowledge or ownership of a critical tool also happens to be a raging asshat who makes the place toxic
|
|
#
?
Jul 28, 2020 16:50
|
|
|
WhyteRyce posted:A common dilemma anywhere is a keeper of tribal knowledge or ownership of a critical tool also happens to be a raging asshat who makes the place toxic Thanks, now we're both disappointed about reality 
|
|
#
?
Jul 28, 2020 17:01
|
|
|
I saw the Murthy news on Yahoo News yesterday before I saw the employee email go out.
|
|
#
?
Jul 28, 2020 17:04
|
|
|
It was leaked to Charlie D. a week before it was supposed to go out so comms was unprepared.
|
|
#
?
Jul 28, 2020 17:15
|
|
|
edit n/m
Josh Lyman fucked around with this message at 08:42 on Jul 29, 2020 |
|
#
?
Jul 28, 2020 17:25
|
|
|
Nomyth posted:While being careful not to cut out exceptions to the rule? Total headchopping isn't really as great of an idea as it seems on paper, especially if documentation and other efforts to make such a transition happen result in a brain drain of what good knowledge there is left. To give you an idea of the level of psychosis in that tier of management, there are department managers with the title of vice president who walk the office area floor every day at certain times, writing down the names of people at their desks ( before covid) to inform their managers and question why they are not attending a particular meeting. The top of the manufacturing org is extremely bad through and through and unless aggressive cuts are made to that tier Intel will never get out of this mess or fix any of it's other problems, most of the high level managers are not interested in success of the process, they only care about not having blame waft in their direction.
|
|
#
?
Jul 28, 2020 20:25
|
|
|
I don't know poo poo about semiconductor manufacturing but I can tell you that you can't fix a broken culture by cutting. You need to bring in people who understand that good culture looks like and then give them the power to teach, model, and enforce that culture. It can be done surprisingly quickly if you're willing to be ruthless about it. In this context, ruthless doesn't mean firing people, it means observing them, critiquing them, and suppressing the bad old culture every single time it rears its ugly head.
|
|
#
?
Jul 28, 2020 22:09
|
|
|
Zorak of Michigan posted:In this context, ruthless doesn't mean firing people, it means observing them, critiquing them, and suppressing the bad old culture every single time it rears its ugly head. It doesn't mean firing everyone, but you gotta replace at least some of the leadership: you need to vacate some spots to be able to slot in those "good culture" people and empower them with enough authority to be able to make their critiques mean something. Otherwise you just get the same ol' leadership doing the same thing, but occasionally having to sit through a review where they're told they need to do X Y and Z and they wave it off because who the gently caress are these guys to tell them how to run their poo poo? Basically, for culture transformation to happen, you either need an absolute genius who can simply convince everyone to want to get on board the good culture train (and these people do exist, but they are exceptionally rare), or you need to convince people that they'd better get their rear end on that train or they're gonna need a new job, which is the less effective but far more common method.
|
|
#
?
Jul 28, 2020 22:16
|
|
|
If I had a nickel for everytime I had to sit through a "this is going to be different and here is why" or "this is a culture change" reorg presentation
|
|
#
?
Jul 28, 2020 22:23
|
|
|
I've interacted a lot with Ann Kelleher in various roles. I respect her a lot, and she's one of the best leaders in the company.
|
|
#
?
Jul 28, 2020 22:30
|
|
|
WhyteRyce posted:If I had a nickel for everytime I had to sit through a "this is going to be different and here is why" or "this is a culture change" reorg presentation Generally those are led by and keep the same people in positions of authority. If you wanna change culture you need to remove anyone responsible for the current culture. Hence why it never actually happens.
|
|
#
?
Jul 28, 2020 22:44
|
|
|
Some Goon posted:Generally those are led by and keep the same people in positions of authority. If you wanna change culture you need to remove anyone responsible for the current culture. Hence why it never actually happens. No these weren't. I've heard it from same people, new people from other departments after a reorg/merge, external hires from other companies. They usually fail for a variety of different reasons from new people being all talk, changes need a few years to fully bake after which the person leaves or is booted and the new person makes their own new changes, to just too many entrenched lifers who resist change and default back to doing things they way they used to Making quick drastic and effective change is really hard for a large, old company to do. And as already mentioned, the more aggressive you are the more operational knowledge you lose. It's a very tall order they have to do. WhyteRyce fucked around with this message at 23:04 on Jul 28, 2020 |
|
#
?
Jul 28, 2020 22:59
|
|
|
WhyteRyce posted:If I had a nickel for everytime I had to sit through a "this is going to be different and here is why" or "this is a culture change" reorg presentation I've gone through a reorg pretty regularly every 2-3 months. It's unreal how many times my second and third-level acronym has changed. It's like people are paid fulltime to move boxes around on an org chart.
|
|
#
?
Jul 28, 2020 22:59
|
|
|
Intel's a large labyrinthine bureaucracy at this point with 100K employees. It's pretty difficult to transform an organization like that, and mass firings or replacing the top layer with outsiders is unlikely to be the answer. I suspect you could parachute in Jensen Huang and he'd make a mess of things as an outsider without history in the company and a group of mid and higher tier experienced staff he could lean on.
|
|
#
?
Jul 29, 2020 00:00
|
|
|
Maybe Jim Keller tried to change things and jumped ship when he realised it was a lost cause.
|
|
#
?
Jul 29, 2020 00:16
|
|
|
It only took nearly 20 years, but Intel finally wore all the tread off of Merom/Conroe.
|
|
#
?
Jul 29, 2020 00:34
|
|
|
Why would Intel have to change. They're making bank, technology leadership doesn't actually matter.
Fame Douglas fucked around with this message at 00:42 on Jul 29, 2020 |
|
#
?
Jul 29, 2020 00:39
|
|
|
|
| # ? Apr 19, 2024 02:41 |
|
|
Fame Douglas posted:Why would Intel have to change. They're making bank, technology leadership doesn't actually matter. In the short term, you're right. But it's not so simple for the long term. On this trajectory they're going to lose large chunks of the datacenter market, which is enormously lucrative. They're fine with mass sales of cheap chips for now because AMD can't buy enough TSMC time to compete toe to toe, but that can and will change if TSMC keeps getting new nodes faster than Intel, and/or Samsung's 8nm and beyond turns out to be competitive. They're not going out of business anytime soon, but they need to figure their poo poo out or they're gonna end up with cratering stock prices.
|
|
#
?
Jul 29, 2020 01:49
|
|