Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,057
|
Post by Dave on Jan 22, 2021 3:08:13 GMT -8
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,057
|
Post by Dave on Jan 22, 2021 3:17:59 GMT -8
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,057
|
Post by Dave on Jan 22, 2021 3:21:35 GMT -8
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,057
|
Post by Dave on Jan 22, 2021 3:28:05 GMT -8
|
|
bud777
fire starter
Posts: 1,352
|
Post by bud777 on Jan 22, 2021 6:31:33 GMT -8
I think that is a step in the right direction for Intel. I worked at Tektronix in Forest Grove. Intel was our neighbor and people who left Tek, ended up at Intel. This was in the days of the 486. Intel has a very Theory-X culture and took pride in its "rough and tumble" management style. Predictably, they were pretty ineffective in most areas that required innovation, with the exception of chip production.Year after year, the improvements in chip manufacturing kept them afloat. I think they are now producing 12 nm chips, 2 generations behind TSMC. a few years ago, they spent 1 billion internally to bring their software teams up to speed, I am not aware of anything that came out of it. They seem to have the right people in place now, but I would watch them carefully. I think they are in worse shape than anyone realizes.
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,057
|
Post by Dave on Jan 22, 2021 6:56:40 GMT -8
AAPL is showing some green in a sea of red (major indexes). Will everyone want to stay in this stock over the weekend or take their profits and run?
|
|
|
Post by Lstream on Jan 22, 2021 7:22:31 GMT -8
I think that is a step in the right direction for Intel. I worked at Tektronix in Forest Grove. Intel was our neighbor and people who left Tek, ended up at Intel. This was in the days of the 486. Intel has a very Theory-X culture and took pride in its "rough and tumble" management style. Predictably, they were pretty ineffective in most areas that required innovation, with the exception of chip production.Year after year, the improvements in chip manufacturing kept them afloat. I think they are now producing 12 nm chips, 2 generations behind TSMC. a few years ago, they spent 1 billion internally to bring their software teams up to speed, I am not aware of anything that came out of it. They seem to have the right people in place now, but I would watch them carefully. I think they are in worse shape than anyone realizes. I am no chip designer, but based upon what I have read, Intel is facing a disruption that was a long time coming, but is finally real. It goes beyond their process challenge vs TSMC. For the first time, Apple has show that it is possible to deliver game changing performance and power efficiency AND successfully deal with software incompatibility at the same time. It is not just a question of x86 Processors catching up. It seems like their underlying CISC architecture CANNOT catch up. It’s like asking a dump truck to catch a Ferrari. Not going to happen. So what are Intel customers supposed to do? Sit around and hope that they are immune from market share loss to Apple, with a permanent impairment that is the x86 architecture? Further, the business model that has been successful “forever” that separates the OS supplier from the hardware has now been shown to be inferior. It feels to me that the PC industry will be forced to move to ARM, in partnership with Microsoft. But Intel is no-where with ARM. They had an ARM business years ago, but sold it to Marvell. This feels like an opportunity for Qualcomm and/or NVidia to ascend at the expense of Intel if they choose to. PC volumes will never approach mobile volumes, so it is not obvious that the business proposition works. Apple has another huge advantage - a common processor platform for everything they make. Bottom line, is that I agree with your bolded comment. I can’t see how they successfully navigate this disruption. Right now they are lucky that the Mac has such small market share. Who knows how long that condition is going to be true. The M1 is a massive “Oh-Shit” moment. Intel had to have known it was coming, and did nothing about it. Now they are in a huge jam.
|
|
SomeJuan
Member
Taking a nap…
Posts: 321
|
Post by SomeJuan on Jan 22, 2021 8:30:43 GMT -8
I think that is a step in the right direction for Intel. I worked at Tektronix in Forest Grove. Intel was our neighbor and people who left Tek, ended up at Intel. This was in the days of the 486. Intel has a very Theory-X culture and took pride in its "rough and tumble" management style. Predictably, they were pretty ineffective in most areas that required innovation, with the exception of chip production.Year after year, the improvements in chip manufacturing kept them afloat. I think they are now producing 12 nm chips, 2 generations behind TSMC. a few years ago, they spent 1 billion internally to bring their software teams up to speed, I am not aware of anything that came out of it. They seem to have the right people in place now, but I would watch them carefully. I think they are in worse shape than anyone realizes. bud777, Intels newly announced chips the 11th Generation Core i5, i7, and i9 are all on 14nm for 2021. Speaking of Intel talking about “the life style company”, Apple. A good read from John Gruber today... daringfireball.net/2021/01/incoming_ceo_of_a_sand-polishing_company_in_oregon_makes_curious_remarkMy parse of this... Intel is in a world of hurt.
|
|
4aapl
Moderator
Posts: 3,598
|
Post by 4aapl on Jan 22, 2021 8:45:53 GMT -8
I think that is a step in the right direction for Intel. I worked at Tektronix in Forest Grove. Intel was our neighbor and people who left Tek, ended up at Intel. This was in the days of the 486. Intel has a very Theory-X culture and took pride in its "rough and tumble" management style. Predictably, they were pretty ineffective in most areas that required innovation, with the exception of chip production.Year after year, the improvements in chip manufacturing kept them afloat. I think they are now producing 12 nm chips, 2 generations behind TSMC. a few years ago, they spent 1 billion internally to bring their software teams up to speed, I am not aware of anything that came out of it. They seem to have the right people in place now, but I would watch them carefully. I think they are in worse shape than anyone realizes. I am no chip designer, but based upon what I have read, Intel is facing a disruption that was a long time coming, but is finally real. It goes beyond their process challenge vs TSMC. For the first time, Apple has show that it is possible to deliver game changing performance and power efficiency AND successfully deal with software incompatibility at the same time. It is not just a question of x86 Processors catching up. It seems like their underlying CISC architecture CANNOT catch up. It’s like asking a dump truck to catch a Ferrari. Not going to happen. So what are Intel customers supposed to do? Sit around and hope that they are immune from market share loss to Apple, with a permanent impairment that is the x86 architecture? Further, the business model that has been successful “forever” that separates the OS supplier from the hardware has now been shown to be inferior. It feels to me that the PC industry will be forced to move to ARM, in partnership with Microsoft. But Intel is no-where with ARM. They had an ARM business years ago, but sold it to Marvell. This feels like an opportunity for Qualcomm and/or NVidia to ascend at the expense of Intel if they choose to. PC volumes will never approach mobile volumes, so it is not obvious that the business proposition works. Apple has another huge advantage - a common processor platform for everything they make. Bottom line, is that I agree with your bolded comment. I can’t see how they successfully navigate this disruption. Right now they are lucky that the Mac has such small market share. Who knows how long that condition is going to be true. The M1 is a massive “Oh-Shit” moment. Intel had to have known it was coming, and did nothing about it. Now they are in a huge jam. I started out in the semiconductor business for a couple years, so have just enough knowledge for this problem to bug me. I think the problem is that this deficit, of being a couple generations behind, can be split into at least 3 areas. The first is the underlying machines, which neither Intel nor TSMC make. You still have to buy them and install them, a huge capital cost, but that's not really a moat as long as you don't pull an Apple and buy out all of the equipment you can get your hands on, as they've done from time to time to keep up with demands rather than as an anticompetitive measure. But the other 2 are the big ones. One is the actual process. One analogy would be race cars, where it's not just the car but also the driver and team. Just because you can throw down tiny details onto a wafer doesn't mean that it's optimized, or even working. Even having briefly been a process engineer here in this space, I initially wasn't putting much weight here. Probably because I wasn't working on cutting edge stuff at the time. And the 3rd is the actual chip design, an important part that I know even less about. Or nothing. But with those things, is there any reason they can't skip a generation? With buying equipment there isn't one. With process engineering and chip design, I imagine that it would be ideal to learn from going through each generation, and that things will be more rocky if you don't, but as an outsider I wonder if it really is impossible. Do you really have to read all the books in the series before reading book 7? What if you read a recap? It might not be quite as fluid, but it's not a showstopper, especially if you bring in some outside experience. Intel has had problems for years, and I remember meeting a guy on the ride around the lake 14 years ago, who worked at Intel in Folsom. Part of that is just the cycle, bringing something cutting edge and complex from not working up to a decent yield. I don't know all of their problems nor their solutions. But Intel is a big company with money to make it through dark times. And telling a bunch of engineers "It can't be done" often gets proven wrong.
|
|
|
Post by Lstream on Jan 22, 2021 9:34:43 GMT -8
I will try to locate an excellent article I read on this a while back. Basically it comes down to a fundamental inefficiency in the CISC architecture. Primarily due to the nature of the instruction set length. In CISC it varies, and is unpredictable. In RISC it is always the same, and can be parallelized better. Skipping a generation can’t solve this architectural handicap, if Intel stays with CISC. This article provides some insight. EDIT - for a long time, Intel could just keep cranking up the clock frequency to deliver more performance. That is no longer a quick and easy upgrade path, considering the thermal challenges this brings.
|
|
4aapl
Moderator
Posts: 3,598
|
Post by 4aapl on Jan 22, 2021 10:17:51 GMT -8
Basically it comes down to a fundamental inefficiency in the CISC architecture. So basically that is like a boat lock. It’s hard to fully utilize with different shapes and sizes, especially if you were limited to only 4 at a time. You can do an ok job, but it's not optimized. There's room for improvement. But it seems like that in itself wouldn’t prevent using a smaller architectural. Am I wrong?
|
|
|
Post by Lstream on Jan 22, 2021 10:34:34 GMT -8
By smaller, do you mean getting to 5 nm or better?
|
|
4aapl
Moderator
Posts: 3,598
|
Post by 4aapl on Jan 22, 2021 10:40:53 GMT -8
By smaller, do you mean getting to 5 nm or better? Moving from their current size to state of the art, whatever size that is. It would (EDIT: add) NOT be optimized, but it should work, right? From a chip design and functionality standpoint, not from a manufacturing and hardware engineering side. (EDIT: looks like cold fingers on the lift for some reason didn't get that NOT in there, which changes the query dramatically)
|
|
|
Post by Lstream on Jan 22, 2021 11:06:25 GMT -8
By smaller, do you mean getting to 5 nm or better? Moving from their current size to state of the art, whatever size that is. It would be optimized, but it should work, right? From a chip design and functionality standpoint, not from a manufacturing and hardware engineering side. They have a buzz saw of simultaneous challenges: 1. Process - they are a couple of generations behind TSMC, right now. And TSMC is not standing still. I don’t know why they are in this spot, so I can’t comment on if and when they can get competitive. 2. CISC vs RISC - the problem with the CISC architectural inferiority can’t be fixed. Intel could build a RISC part, but they are not a software company. Meaning, how do they pull off something like Rosetta to enable a smooth transition? The whole Universally Binary thing is above my pay grade, but I suspect there are challenges there too. 3. Apple has finally gotten to the point of showing that almost full vertical integration including hardware and software can deliver a better experience, at a competitive cost. Next up, Apple will do their own radio. Who knows what that means. Maybe Intel and it’s partners can deliver “good enough”, but Intel is looking at a situation where the fundamental business model is no longer the best game in town. 4. Focus - Apple has basically one platform to support. So they can optimize the hell out of it. Intel is a merchant chip supplier with different masters. Execution speed and effectiveness are impacted by this. Intel has one advantage in that they supply the dominant PC OS with Windows. That will buy them some forgiveness if they can’t deliver something that is just as good as the M1. Guess we will see how much that advantage buys them. I think they are years away from having an effective answer to the ARM challenge. Which is an eternity in the tech business.
|
|
|
Post by archibaldtuttle on Jan 22, 2021 11:23:12 GMT -8
Be careful not to start any car talk, people!
|
|
|
Post by Lstream on Jan 22, 2021 11:47:09 GMT -8
Be careful not to start any car talk, people! I might want an e-bike. Is that off-limits too?
|
|
4aapl
Moderator
Posts: 3,598
|
Post by 4aapl on Jan 22, 2021 12:27:21 GMT -8
2. CISC vs RISC - the problem with the CISC architectural inferiority can’t be fixed. Intel could build a RISC part, but they are not a software company. Meaning, how do they pull off something like Rosetta to enable a smooth transition? The whole Universally Binary thing is above my pay grade, but I suspect there are challenges there too. This stuff isn't that complex. There's nothing that fancy about Rosetta. It's just a translator, making things backwards compatible. It helps to have a fast processor to keep things moving, but really you are dealing with software originally designed to run on a slower processor, so taking a speed hit from the current state of the art can still give you a pretty fast translation. And universal binaries are just like getting an instruction manual in both English and Spanish. It's a waste of space if you only ever use one of the choices, but it has both so it's ready no matter which language you use. While I think I'd love to see Apple take over the computing space (there can always be unintended consequences), I'm just not convinced that even being double the speed of the competition is going to get that huge of a percentage to switch. Look at iPhones. They knock the socks off of most of the competition in speed tests, and I believe are even considerably faster than the best the competition has. But for most people, what does it matter? I'd imagine that Apple gets more switchers from having the complete package that just works well, than the actual exact speed. Heck, the freaking colors probably get more people than the exact speed. What about cars? Or for archibaldtuttle, what about boats? Just because a more powerful "boat" exists, it doesn't make everything else obsolete. Not everyone switches. Some percentage of people do. But many don't, and not just because of a variety of functionality (speedboat vs wake board boat vs ski boat vs fishing boat vs sail boat vs dingy vs kayak vs paddleboard). There's cost. There's brand loyalty. And there is design. Apple often focuses on a few products so much that it doesn't give all the choices you may want. No wake board boat for you! The high end users may be more flexible if they have uses that really need the power. And they can justify the cost. Like a monstrous $1400 pro chainsaw, instead of a $100 economy model for the do-it-yourselfer with one small tree to deal with. But there's a lot of those homeowners, and not many of them are interested in slabbing big trees. There's middle ground there for those that are processing a lot of firewood each year, but most people just don't have the need if all they are doing is some aggressive pruning and taking care of an occasional small downed tree. We'll see. But it's never a good idea to downplay the competition too much, especially one who is still the dominant PC processor designer and manufacturer, and has been through the ups and downs before. Maybe they won't make it through this time, but it seems likely that they will at least in some form.
|
|
chinacat
Moderator
AAPL Long since 2006
Posts: 4,425
|
Post by chinacat on Jan 22, 2021 12:38:03 GMT -8
AAPL is showing some green in a sea of red (major indexes). Will everyone want to stay in this stock over the weekend or take their profits and run? I would guess that most folks expect a very good earnings report next week, so...
|
|
|
Post by dc930 on Jan 22, 2021 12:38:39 GMT -8
Cheers!
|
|
|
Post by artman1033 on Jan 22, 2021 12:46:21 GMT -8
AAPL ALL TIME HIGH! $139.85 All Time Highest TODAY intraday
|
|
|
Post by Lstream on Jan 22, 2021 12:46:25 GMT -8
I am not predicting the end of Intel, just agreeing with a poster who said they are in world of hurt, and that many do not fully understand how much. I am also not predicting a large increase in Mac market share. On that, I fully get that all kinds of people will never move away from Windows, no matter what.
|
|
|
Post by artman1033 on Jan 22, 2021 12:46:26 GMT -8
AAPL ALL TIME HIGH!$139.85All Time Highest TODAY intraday 104,668,597
shares traded today +$2.34 TRILLION
|
|
benoir
fire starter
*
Posts: 1,314
|
Post by benoir on Jan 22, 2021 14:32:21 GMT -8
Any ‘wheeeeeeeeee’s out there today?
Edit.... for some reason I often double post on my iPad. Posting when I’m half asleep perhaps 🤔....
|
|
benoir
fire starter
*
Posts: 1,314
|
Post by benoir on Jan 22, 2021 14:34:36 GMT -8
Any ‘wheeeeeeeeee’s out there today? (I’m not sure any more whether we do that for big gains and ATH’s or for when Artman’s post’s are bold and green?!)
|
|
JDSoCal
Member
Aspiring oligarch
Posts: 4,181
|
Post by JDSoCal on Jan 22, 2021 15:02:02 GMT -8
Great day.
|
|
4aapl
Moderator
Posts: 3,598
|
Post by 4aapl on Jan 22, 2021 15:05:33 GMT -8
Any ‘wheeeeeeeeee’s out there today? (I’m not sure any more whether we do that for big gains and ATH’s or for when Artman’s post’s are bold and green?!) Wheeeeee! Even though by absolute gainage it might not have been the biggest day, it was pretty satisfying to come out of the cellar and into the green, setting a new ATH while the S&P was down just a hair. RSI is still sub-70. This time is different, as they always are, but I'd hope that AAPL doesn't gain so much this time pre-earnings that it doesn't keep up the momentum even after great earnings. I know I know, JD's maxim of enjoying the moment. Thanks everyone, for all of your posts! And before I forget, there's at least a few people out there with accounts that have yet to post, that I've seen popping up on the Poll or the Visiting list. It would be great to hear from you guys, maybe on something like what your first Mac or iPhone was? Or when or why you bought your first shares of AAPL? These are great times for Apple and AAPL. Join on in!
|
|
|
Post by nwjade on Jan 22, 2021 17:24:42 GMT -8
Any ‘wheeeeeeeeee’s out there today? (I’m not sure any more whether we do that for big gains and ATH’s or for when Artman’s post’s are bold and green?!) Wheeeeee! Even though by absolute gainage it might not have been the biggest day, it was pretty satisfying to come out of the cellar and into the green, setting a new ATH while the S&P was down just a hair. RSI is still sub-70. This time is different, as they always are, but I'd hope that AAPL doesn't gain so much this time pre-earnings that it doesn't keep up the momentum even after great earnings. I know I know, JD's maxim of enjoying the moment. Thanks everyone, for all of your posts! And before I forget, there's at least a few people out there with accounts that have yet to post, that I've seen popping up on the Poll or the Visiting list. It would be great to hear from you guys, maybe on something like what your first Mac or iPhone was? Or when or why you bought your first shares of AAPL? These are great times for Apple and AAPL. Join on in! So here we are finally back at $139.07 essentially the close on 01/03/19... Oh wait, that was pre 4:1 split and we're up 4X since then, never mind I remember vividly CNBC panels of numbnuts debating the sh*t out of whether aapl was a buy at $140-$150 back in that time frame, unreal.
So here we are two years later sitting at an ATH of $139.07 with a RSI of 65.72.
Trailing PE at 42.40 is high but knock on wood, excellent Q1 earnings will reel that in and the aapl party will continue on into Q2. I'm really enjoying the moment! Cheers to the longs!!!
|
|
SomeJuan
Member
Taking a nap…
Posts: 321
|
Post by SomeJuan on Jan 22, 2021 18:57:45 GMT -8
Ik,ik,
I said i am retired, not tired though.
I spoke to my uncle tonite who spent 23 years working as a chip analyst doing process verification of CPU’s after the wafers were “packaged” on AMAT devices, then his team did analysis and inspection on each CPU for testing purposes, he said the term is called “binning” now. Pass/Fail/Suboptimal but salable at discount. He was a QA manager. He says yields at startup could be as low as 7-8% in first runs and 50-60% in pre-production, where they would seed cpu’s to OEM’s, and full on production of 90% +/-. Failures were universally 7-8% on full on production, and 2-3% were “binned suboptimal, but salable at lower clock speeds”.
He broke his QC bones at IBM in East Fishkill, NY working for IBM, doing QC on IBM G4 processors for AAPL, at 450-500 Ghz! Also Using AMAT products for production, and QC devices.
He worked on 200mm wafers, then the newer 300mm. These were 140nm SOC’s.
He is reading these posts... 😬
He said Lstream is a smart sun of a bish. I agree.
He wants me and him to co-write a sub 2000 word article that explains the 5-6 step process of what a fab does, i am going to... is that too long for this place?
He says it wont be shedding a lot of good light on Intel... and does not want his name on it, i can see his reasons, he still get’s benefits from Intel. He is retired, not a retread!
Yes/No?
|
|
4aapl
Moderator
Posts: 3,598
|
Post by 4aapl on Jan 22, 2021 20:07:30 GMT -8
He wants me and him to co-write a sub 2000 word article that explains the 5-6 step process of what a fab does, i am going to... is that too long for this place? He says it wont be shedding a lot of good light on Intel... and does not want his name on it, i can see his reasons, he still get’s benefits from Intel. He is retired, not a retread! Sounds great. We'll find a place for it. Not in the daily thread, but maybe a shared knowledge folder or something. We already try to share somewhat related knowledge at times, directly or with references. For instance, book and movie reviews, with some of the media on topic and others not. (Just finished "The Happiness Myth". I wasn't as in to the 2nd half, but the first half did make me rethink some preconceived notions I had) I know some here are interested in learning more about semiconductor processing, and the pitfalls. I had a book on it, maybe given to all of the new hires in our group at Motorola. We did 6 month rotations in different areas, so we learned a bit about a bunch of stuff. My first was in photolithography, but that was on 1970's machines, dealing with power mosfets. Not really cutting edge. But Ion Implant and Chemical Mechanical Polishing were much more cutting edge, all as a Process Engineer. We'll find a place, and I know a bunch of people here will be interested. Thanks macentropist.
|
|
|
Post by Lstream on Jan 22, 2021 20:08:10 GMT -8
I would read that fab article in a flash. Please do.
|
|