Samsung Electronics cultural issues causing disasters in foundry, LSI, DRAM
semianalysis.substack.com
semianalysis.substack.com
Under such a restrictive and top-down culture like this, I believe nothing interesting can emerge. I can only expect the downfall of the company. Nowadays I'm trying to work as little as possible and desperately searching for other jobs.
What the hell kind of mentality is this?
I’ve been in a workplace that cracked down on phone use. Everyone was pissed. I considered carrying a small block of wood painted black in act of subversion.
sighs Traditional Asian mentality. I, and definitely many others, hate it, but it's definitely not just Samsung.
Samsung takes that cheating to 11 and there is no way to turn it off. It’s not “store mode”. It’s permanently “enhanced” to the point of absurdity.
They just don’t care enough to have two modes.
Not to defend Samsung, but this isn't such an objectively straightforward task as you make it sound. It's impossible, in the general case, to convert colors from one colorspace to another without information loss. So there's an element of subjectivity and judgement in selecting the algorithm used. In other words, they're all "enhanced", and your complaint is simply that Samsung has poor taste.
Of course this all assumes that the display is doing its own color management. This is exactly what you want with a standalone TV, but for a computer display you really want to just provide the computer with the ICC profile and give it the lowest-level access to the pixel values possible, so that the user can assume control over the rendering intent - for that you'd want some kind of "direct mode". It's quite right to criticize Samsung if they do not offer such a mode, but are the other brands any better? The trend for TVs seems to lead away from being good general purpose displays, and towards being standalone devices.
Even in the general case that's not true for new televisions: These typically convert from a smaller space to a large one, which can be done 100% losslessly and accurately as intended.
This is quite common, because as TV panel capabilities have largely outstripped the distribution and encoding standards. For example, Rec.709 (the HD standard) is smaller than what any modern 4K TV can display.
The 4K Rec.2020 standard is huge, but it only exceeds typical panel capabilities along the green axis. This may seem like conversion would be lossy, but if you look carefully at video metadata, it ofen specifies that the "content gamut" is smaller. That is, most modern 4K HDR content is mastered on a display that "merely" has a gamut like Display P3, so that's all you need to reproduce it 1:1.
None of this is done by Samsung. They always stretch colours up to the maximum capability of the panel. These days, that's very close to the full Rec.2020 gamut and looks downright garish. Even if "AI enhanced" or whatever, it's wrong. Colours are distorted and nowhere near the original intent.
Source: I work an audio-related mobile app. The audio engine code is littered with comments that read like "On phone X, Y happens and we do Z to work around it". In about 75% of the cases X is some Samsung model. And before you ask, X is the phone it was discovered on so there's no guarantee that the same issues don't occur on other phones but somehow Samsung is the name that always pops up.
On a side note, My favorite phone I've ever had was an LG Envy 3. It was about the size of the palm of my hand and, closed, had a 1.25" screen and a standard 9-key keypad. It could also flip open to reveal a full qwerty keyboard and a 3" screen with some respectable stereo speakers.
I'd like to believe the process has to be at least a little bit similar to how agile, innovative and leading companies in the West can within a decade or two turn into stiff, burocratic slow moving behemoths that at best manage to buy any upcoming competition instead of innovating on its own.
i'd say that how SK was structured after the war allowed them to flourish, basically a dictator appointing diff people into diff industries and having a state-sponsored acceleration. add to that a drive to succeed and fight and you can somewhat see the building blocks of their success. how SK has progressed is a miracle story.
now that they are seeing the fruits of their labor my guess is you have the "parasites" coming up trying to jockey for position. or with chaebols (these big family dynasties that own a lot of industry in SK) they have their kids/family members who may or may not be competent taking over. essentially what made something like samsung takes a lot to sustain into the next generation.
Why are you so desperate to use your phone?
This is practice in many companies and I hear at Apple it is even more restrictive at certain places.
Not able to use phone I never heard or experienced. We regularly test out apps in phone, test out devices (non test ones), we test competition ones, etc. And when one can get on a Zoom or Team's call, what use is restricting phone usage.
About 12 years ago, I consulted, on-site, for about a year with a Japanese company with a very very very untrusting corporate culture. (I was consulting on electronics and firmware.)
Every USB port on every computer was super-glued shut (yes, I know...) And literally 100% of the internet traffic went over a T1 (for those of you just out of diapers, that is a 1.5 Mbps line) to Tokyo and back. Corporate saw every bit of internet activity.
This company, when I talked with the VP of engineering, told me "we have flexible work hours; you can take lunch 12:00-12:30, or 12:30-1:00pm. As long as you're back at your desk within 5 minutes of lunch end, we will not be required to make note of it"
Man that was a horrible gig. But I made enough $$$ to overcome the agony. I have so many consulting stories... my daughters keep telling me I need to write a book.
"When Bill and Michelle ruined a conference room table"
"When Jayanthi [no I'm not changing her name] had a meltdown and threw a sharp object at the head of a subordinate"
"When Tom impregnated his secretary during his divorce, and was basically forced to marry her the day after his divorce was finalized"
"When I put a choke hold on a corporate security guard who incorrectly identified a computer peripheral in my backpack as confidential company property"
I think the 100 people who bought the book would really enjoy it
Aah yes. The sign of entirely level-headed and well-managed companies is when they blame the fact that they can't 996 their skilled workers.
The legislation aspect is really concerning but not surprising as chaebols have an iron grip on the Korean economy.
Tough times for Samsung, but in a lot of ways, you yield (pun intended) what you sow.
"We just need to copy Korea" is not a good narrative if you want to keep high-quality American engineers interested in a long term career path with you.
I lasted barely 3 years before the lack of volition got to me. Went to contract for a video game developer afterwards (huge pay cut) just to clear my mind and reconnect with reality.
Damn, that's really saying something about your former employer.
To put things into perspective, not many people ever get to set foot inside of a leading-edge fab. I've been inside multiple, both in America & Korea. I've been able to sit in meetings with engineers across the entire spectrum and participate in multi-national system upgrade efforts.
The best way to describe the whole experience is like working inside a starship. Genuinely, it feels incredible to walk into that room and see billions of dollars of the most sophisticated hardware on earth all working together in relentless harmony. Simply sitting in the engineering offices and seeing the real-time logs scrolling gives you a sense for the monstrosity just a few firewalls over. The techniques & hardware are exotic by default everywhere. It almost never gets old.
But, just like all good things this experience fades with time. I wanted to build new amazing things and being one mere engineering pleb inside this gigantic organization makes that a difficult gambit. Perhaps if the organization was willing to explore more experimental / "internal startup" style work, I could be compelled to review future opportunities. There were plenty of problems to solve but getting a design meeting or a piece of IT infra to run something on were nearly impossible during my time there.
Asking as someone who has 0 experience with either, but this is how I’ve heard Apple works too? If so, why is Apple doing well while Samsung is not?
Take Steve Jobs. He was infamous for his perfectionism. But on the other hand he trusted his team. When several of his execs insisted that they should build an App Store, which he initially opposed, he folded and let them do it. He once said there’s no point hiring A grade engineers and then telling them what to do and how to do it. I hire A grade engineers so they can tell me.
(0) https://donmelton.com/archives/ (1) https://youtu.be/N8Vz1BeymHE
I had a friend that was interviewing with Apple. Apparently the orgs are so siloed that the teams that wanted him didn’t know about each other. It appears to be a very rigid sort of organization.
Allow me to add some context. The Samsung culture could be described as militaristic, with accompanying rank and ceremony, therefore I can see how you identified the top-down chain-of-command relationships.
[edited]
I think most people would accept that Intel has/had similar problems based in financial optimization, but that doesn't directly reflect on 5000 years of western culture.
There hasn't even been 5000 years of western culture...
But if you go back to the Early Greek era of the Minoan civilizations then it is about 5000 years.
It's all slow, gradual accretion though, so you can draw the line in lots of places.
Either way, on any of those benchmarks Mesopotamia precedes it by at least a few centuries or more, though Minoan does get you back pretty far. I don't know the migration patterns, but I think there was influence from Mesopotamia.
Classical Greece ~2500 years is probably often where the line is drawn because that's where some of the foundational works of culture, philosophy, and math had their start and became founding members of the "Western Canon". For millenia, students were trained in Greek and Latin and read the great works of those time periods. But there was a civilization that gradually evolved into that time period, complete with it's own Dark Ages before things climbed back up to what became Classical.
https://en.wikipedia.org/wiki/History_of_Indigenous_Australi...
Yes, the glorious Western civilization did the same pretty much everywhere, but Australian Aboriginal people's culture is unique due to their very long period of isolation - it's a gem and wonder worth studying and learning from. Yet, the Western instinct was to trample on and destroy it, systematically, irreversibly, just for the sake of it. Reading the Wiki page you linked some time back literally left me in tears. How could we. I don't even.
I'd like to add that it's never too late to try to preserve these cultures, traditions and languages. Getting interested to the ones of your place of residence, even if you're not from there originally, can be a great experience.
One of the best ways to do it is through singing and dancing (they go great together), and more generally, there are often chorals and various groups dedicated to these cultures, traditions and languages.
And indeed when you get into this kind of thing, you realize how much culture has been erased (in my case, the Occitan culture), but also you can see how much is left, and how important it is to preserve it.
It also helps a lot when reflecting on our own modern, imposed, culture, and realize that some of the things we think have always been there, have not.
Co-opted first by Alexander, then Caesar. Later Napoleon also dropped by ...
Seems to have slipped everybody's mind here, strangely.
This is like claiming the Islamic Republic of Iran is the same culture as ancient Persia.
That a preindustrial society existed for 5,000 years with very little change has no bearing at all on how a multinational business ought to be run in a modern society. In fact, I would argue that technological advances became possible precisely because we blew apart many of these entrenched power structures.
I am sure that what you wrote is true from your perspective and have therefor edited my comment so as to not produce discomfort.
The big reason for that is any overseas research center like SRA (Samsung Research America) works on future, whereas Korea works on current. When working on future features or tech, it is all about bets and experimentation. And these are wont to fail or be wrong quite frequently.
I don't really know what's more absurd, that Samsung managed to fall behind these guys or that MediaTek actually managed to get a grip on quality well enough to rival and surpass Samsung, a multi billion dollar conglomerate.
In [2], I'm seeing sensors (not made by Mediatek), GPU drivers (Mali, ask arm on that one), and RF (usually regulatory; legalism?), and whatever else is in that textdump. Is there anything that actually does an analysis? Even the blog source eventually just said they didn't know.
His mistakes weren't seen clearly until almost 2019, when it became obvious 10nm was a disaster, and 14nm++++++++ was their only workable product.
We're still seeing it today as Intel has 250w TDP products competing with AMD 140w TDP products.
Semiconductors is a space where mistakes take a while to bit you.
Although as we've seen with AMD, it is possible to come back from them. With the right course correction, Samsung is diversified enough to tough things out. Didn't they have a massive turnaround sometime in the 80's too?
See: Pentium 4
Seems not that far-fetched for Samsung to be facing cultural issues as well and that a company can be successful despite them for a long time. It seems like it's finally catching up to them as their competition becomes stronger.
It's clear that Samsung culture prevents people from admitting mistakes, which of course only leads to more of them as stress builds up.
I remember the Samsung Galaxy Note battery problems, and how apparently it was clear to people outside Samsung what the problem was (too tight a clearance a the corners in the case for the batteries, causing bending of the internal electrodes, leading to short-circuits) before Samsung themselves recognized it.
You can't always be at the top, a sign of good culture is what happens when you slip.
Samsung keeps slipping.
Jesus H. Christ on a fucking hockey stick. I doubt I've ever read such a breathless hit piece with so many unsupported assertions summarily presented as fact.
I worked for Samsung Semiconductor for ten years and left as a software engineering manager. None of these assertions and accusations in the comments here or below leveled at the people with whom I used to work ring true from my perspective.
Communication issues? Not admitting mistakes? Please... enlighten me. Tell me how we worked.
What if I threw in a brisket from Franklin's and told you I would be your pal
Maybe it's an IDM mindset issue? It seems the market has moved on to pure play foundries. For example, Nvidia's first fab partner was STMicro, an IDM, and Nvidia noted STMicro was unable to focus on being a fab partner. After that, Nvidia moved onto TSMC, also working with IBM, UMC, and Samsung at later points in time.
Samsung, from my perspective, mostly focuses on this year's production and developing next year's model. Yes, there is work done planning for new fabs and global supply chain systems, but 98% of cycles are spent on this year's production and designing next year's model.
Samsung accounts for a huge percentage of South Korea's GDP, about 20%. Therefore, Samsung needs to favor stability over taking risks associated with living at the bleeding-edge of innovation. Samsung always did play the safe game when I worked there (and for good reason when you're babysitting a $20 billion factory with $100 million of product that could be scrapped at any time).
Again, I wasn't in LSI but this is my perception overall having worked there, traveled to Korea and learned Korean culture.
I worked at SAS in lots of different areas—Metro, Photo, and PIE (process integration).
The emotional defense above notwithstanding (which I'd characterize as not actually being supported, either...), SAS indeed has very, very deep cultural problems that make for a shockingly bad example of how to do engineering in a large organization. There are several factors contributing to this.
SAS doesn't treat information systems, or things IT-related generally, as an engineering concern. If it doesn't look like a materials science problem, SAS doesn't have the capacity to critically evaluate its impact on rates of production and yield.
Problems, though, range from lots of things, like:
- far too many things in-fab being handled with pen and paper
- reports that have to be tediously assembled manually, so they get sent out once a day, while realistically they should be available to an engineer at any given time (behind a button that generates them on the fly instead of throwing $45k/year technicians at the problem like a 1960s-era secretary pool)
- workers committed to this state of affairs remaining so indefinitely
- Conway-style balkanization and groups who have no idea what their role is ("are they asking me or are they telling me?")
- a bunch of dumb macho shit that is used to prevent critical productivity problems from being addressed because being lazy and not addressing it is somehow virtuous and not lazy—it's the application of basic engineering principles to eliminate a problem that's lazy!
- an effect that I have no way to describe except as the proliferation of mid-40s to 50s men of mediocre talent (to a degree that's even worse than what you can expect to see elsewhere, generally)
- an overreliance on the labor pool of people with military experience / government work for no reason other than they've been conditioned to survive in the kind of bureaucratic senselessness that SAS exemplifies
- unabashed credentialism, which gets people hired who shouldn't be
- Microsoft Office products being abused for everything—including areas where a different Microsoft Office product would be the most appropriate tool for the job (reaching for PowerPoint when the problem calls for Word—or any word processor)
- work showstoppers caused by truly ridiculous problems, like Windows crashing a half a dozen times a day; Caps Lock on a fab tool keeping people from remoting in from within the office, so they have to suit up, go into the fab, press the Caps Lock key, and return to their desk and resume work; lots of monkeying around just to share a file with someone, because everything has to be ferried through a proprietary document sharing system that amounts to a crippled, quasi-web-based thing that wants to be Windows Explorer-cum-Google Drive
- unit parts (i.e. manufacturing divisions) basically re-discovering/re-inventing the value version of control and implementing it in an ad hoc way for the tools that they ineract with, poorly—not to mention all opportunities for human error (which does happen regularly but no one does anything about it because 40% of the job is CYA and the CYA approach tends to be "don't let anyone find out this happened")
- tons of janky-ass software from S1, like "Simax", which is written in a proprietary Lisp-ish(?) that uses ActiveX to run in an IE9-10-11 window and that re-invents its own window management and basic form controls, was never fully localized into English (so you're clicking buttons with terse messages in Hangul that still don't make sense even if understanding the language is no problem), and times out and ends your session if you look away from it for 10 minutes (or whatever)
- overall just really bad engineering discipline; people putting wafers on hold and not being clear about they're expecting to do with them next or what they need someone else to do; people inventing new acronyms all over the place (because job security and also "mediocre talents") and using them liberally in the notes they do leave for others; people not documenting why something happened, which also happens to help when covering something up and hoping that no one has the time budget to untangle things to the point where it's obvious/incontrovertible that something did go wrong; also lots of people not having a dedicated workstation of their own, so when shifts change they have to be out of their seat so it can be ready—no putting in a little extra time to make sure that the kinds of fires that were put out that day won't just happen again tomorrow and the day after, ad infinitum
In a massive operation running 24/7 that never stops to catch its breath, it's chaotic. There are certain techniques I adopted that helped me be successful there, but I definitely understand your frustrations.
No, it isn't. Describing it like this is needless apologetics.
The "chaos" is not even necessarily the worst part of the problem. It's these kinds of just-so dismissals that have the effect of framing the chaos as somehow unavoidable—the response that goes, roughly, you'll learn; this a natural consequence of what we're doing; it's a big operation, and this is what it looks like when you play with the big boys. I can tell you: I am a big boy. I have too much experience to the contrary for these casting couch excuses to work. Much of the day-to-day toil at Samsung is inexcusable and entirely avoidable. There is no excuse, for example, for dealing with version control in a billion dollar manufacturing operation like we're living in 1995 or a neverending sophomore-level group project.
I'd never looked at it that way, and it succinctly describes why the least devops-mature organizations I've worked with have been those that heavily penalize any admission of failure. And vice versus.
Unfortunately, without active checks and balances, orgs will tend to the blame game. Avoiding it requires an org-wide commitment to openness around failures.
I've always been surprised when I see things like this, from websites with all the controls reinvented to MDI-based Windows apps. Where do they get the _time_ and designer resources to waste all their time doing this?
Maybe I just learned programming in a different way, since one of the first things I learned was "laziness is a virtue".
It's finally gotten to the point where there is the action potential to fix it in my case, but for a lot of systems with a complex and expedient lineage like this, it never gets there. So garden well, I guess, is the takeaway.
No one ever sat down and said "How can I design a terrible system that's a pain in the ass to work with and unable to be extended?"
Everyone sat down and said "How can I make that one thing that's actually my job faster?"
Never ascribe to malice, that which can be described by laziness, etc.
It seems like it would take more work to specifically design a system to ensure job security than to just haphazardly design without coherent architecture, and consequently be the only one who understands it.
As a user, interfacing with these systems was always hell. Setting up online banking on a new computer would take the better part of a day, and would fail if you didn't have IE security exceptions set, if you were missing Korean font packs, or if your name was too long. And different ActiveX controls installs were required by every webapp. Our company (healthcare) finally rolled out a native Windows app in 2019. It includes its own floating tile manager, and for tasks like viewing PDFs, browser frames are now embedded in a tile...
One guy I know works 6 days a week, Monday-Saturday. He goes in on work around 2pm Sunday to get caught up. This isn't the 'crunch' schedule. It's just his regular work schedule.
I’m not sure if this is supposed to be a satire or not. The forcefull and quite frankly agressive way you denied ever having a problem with admiting mistakes makes me think that there might be something to this allegiation.
Every company I ever worked with had occasional issues with communication, and even rarer issues with admiting mistakes. Large companies are also very uneven. Maybe the parts you worked in were great while the parts these other folks talk about were not so great? Either way swearing and writing in an unprofessional way is not the right way to go if you want to convince people about what you are saying.
It's more an instance of "place where one has worked". Someone in that position at SAS simply doesn't have the experience to confirm or deny the sorts of things the article gets into. Not to equate being a software engineering manager with being a janitor, but it is, on some level, like asking a janitor what was really going on at Enron just because they were in the same building.
Respectfully, that is a straw man. I never made that claim.
I am saying that GGP's claim is unsupported and therefore baseless. Knowing myself exactly how things really work there, I read it as ignorant hyperbole.
To answer your questions, are there problems in Samsung? Samsung has the same problems with communication and admitting mistakes as other companies for whom I've worked of similar size. I've actually seen much worse at small companies, i.e. startups.
I personally never had issues at Samsung. I could directly communicate with anyone in the company and did so, often. When necessary, we coordinated work ahead of time, even after hours. It wasn't a problem.
I don't want to make it sound like working at Samsung was a walk in the park, because it wasn't. Samsung was by far the hardest job I've ever had. But working there taught me that with the right team and good teamwork, you can accomplish anything. We did some amazing things there.
This is some amazing mental gymnastics! You should get into politics!
Maybe Samsung does not have a problem about admitting mistakes. But a problem about admitting mistakes in their problem admitting mistakes procedures. It's always the meta-problems that get you!
Again, this above comment shows people can twist anything into anything and will just stick to what confirms their bias.
What else could there be?? Do you think the company will release this information in a press release??? Do you think people will put their names to these criticisms??
That turned out not to be the cause. https://news.samsung.com/global/infographic-galaxy-note7-wha...
> The negative electrode was deflected within the upper-right corner of the battery
Maybe better clearance would have prevented that from causing a short?
Is this true? That model would NOT work for semiconductors, where each fab generation is a bit of a hit or miss, but the game requires that you gamble.
But the problem isn't internal communication though. It was Samsung Foundry's "sales and marketing" over promise and under deliver. To give them credit they were throwing insane amount of money trying to compete with TSMC. People on HN / Reddit / SemiWiki or whatever forum keep asking why TSMC isn't raising price. Well you should thank Samsung for that. ( But they are also the ones who HN / Reddit / Internet accuse them of NAND / DRAM cartel, guess where those profits went? ).
I also think the DRAM and NAND concern are overblown. It has been well known Samsung took the bet to move forward with EUV earlier than other industry vendors aka SK Hynix and Micron. It is not like Micron is going to go straight to EUV and enjoy the advantage over Samsung. Samsung are also not moving forward with EUV on their NAND product. The only concern would be if Micron decided to increase their capacity and build new Fabs. Otherwise, just like any other commodities ( Corn, Steel, Lumber, whatever ), they are Supply and Demand limited. So from a technical perspective they may be losing, from a business perspective things are still within control. In case anyone wants to compare this to Intel's 14nm, the two are extremely different and their market operate in a completely different manner. Hence why TSMC's founder and ex-CEO Morris Cheung famously said he will never enter NAND and DRAM market.
(Our fridge is quiet, for what it’s worth.)
There is basically no warranty on replacement parts. We stopped replacing that panel, and now we just play the guessing game. Fortunately the fridge does remain functional as a cold box, at least. There are a few other problems, like accumulating ice in the tray at the bottom, but this is not really a Samsung-specific problem, lots of fridges develop this issue.
I have a totally different fridge, and it has a heater at the bottom... but I needed to physically bend the heater a smidge so it was close enough to the drain to keep the drain clear and prevent ice build up. Might help if there's something similar.
Anyway, we also have a samsung washer dryer pair. We had some issues during initial installation, and had to deal with samsung support. It was so bad that we now regret the fridge purchase (it’s probably a different division, but for how long?), and will never buy samsung again. Not sure what brand of SSD to use moving forward.
In related news, I read yesterday that they’re being sued because the tech support chat people (calling them is completely futile) are paid $0 + sales commissions.
This one fact explains more about my interaction with samsung tech support than any article about corporate culture ever could.
(I also know people that worked there. It sounds horrible from the inside too. On top of that, appliance parts suppliers in my area confirm that samsung is one of the worst possible choices for after sales support).
Nowadays it's a running joke among my friends about how much I advocate against buying any Samsung devices to anyone who will listen. I've probably saved 5 people from buying a Samsung TV - instead convincing them to get one of LGs excellent TVs.
Strangely, it's actually a really good fridge and it's one of the only items in our house that DOESN'T connect to the internet.
The LG we already had would occasionally freeze vegetables, the Fisher & Paykel had an internal light which stopped working... and the Samsung seems to be a simple fridge that works well (so far).
Genuinely curious here how you were not only able to verify Samsung refrigerator ownership status for 100K people on Facebook but also that 99% of them were slandering Samsung's good name purely for racist/nationalist reasons.
Unless there's some kind of evidence to the contrary, I would just assume these people are unhappy with the refrigerator they bought.
Fridge: the coils in the fridge compartment freeze up, which is expected because, hey, thermodynamic laws. Hence the heater coil to melt the ice. Which would be fine, except that the little metal tab that is supposed to conduct heat into the drainpipe is too short, so the ice doesn’t melt, thus leaving the drain all plugged up, causing the ice water to overflow, eventually spilling out the fridge door and destroying your cork floor.
Dishwasher: fffuuuuuu… P.o.S. Piss-poor job of cleaning. Does a shit job of filtering food particles out, so they end up in the rinse water, and so end up crudding-up any contact points between dishes. I end up washing “clean” dishes by hand Every. GD. Time.
Oven: self-clean feature cooks the convection fan to death. Should be called self-destructive, not self-cleaning. Thankfully, the induction stovetop continues to work… for the time being.
I will never, ever purchase a Samsung product again.
I would really think twice before I bought another Samsung consumer product.
When you press the power button and it goes to art mode (effectively a screensaver) instead of turning off, most people's next step would be to try holding the power button. Barring that, the manual clearly describes these two different operations if the power button.
You would do well to work in tech support for a few months because this claim is absolutely a complete fantasy.
And now I just discovered my Apple TV actually has a control panel. You don't need to go two layers deep in Settings to put it to sleep. It never occurred to me to a long press on the home button would do anything. (Despite knowing long pressing on the menu taking you to the main menu.)
You need to keep your finger pressing the off button on the remote until the whole thing turns off.
> These various units are allegedly playing the blame game with each other. Samsung LSI (design) is blaming Samsung Foundry, while Samsung Mobile is blaming S.LSI.
The link in this paragraph doesn’t actually have anything about them blaming each other.
> that the foundry is even allegedly lying about the yields
The link here also doesn’t say anything about “lying”. It says they will do a diagnosis/audit around the low yield problem.
I didn’t bother clicking on all the links. But some links are right but kinda erodes the trust in the article when this is happening.
Someone reading that article might have walked away with the impression that Samsung's collapse is imminent, which seems to be the narrative often touted on the internet circles by people with some type of stake in TSMC or its allied subsidiaries.
"Clients and employees of SemiAnalysis may hold positions in companies referenced in this article."
I mean from the authors disclaimer here we can see why the article reads like some Zerohedge article. I hope the author can afford expensive lawyers because there's just no way Samsung is going to let the author off the hook, they've gone after journalists for way less.
You have a longer chain of communication between you and them, so things get muddied.
I imagine that foundry news written in Chinese/Korean/Cantonese is much more well written and in depth.
Which leads to the narrative equivalent of a PowerPoint presentation where every slide is just plastered with numbers from corner to corner.
Somewhat surprisingly, best writings, at least at popular level, come from Japan. Japan is the highest literacy society on Earth (if you look at OECD PIAAC, there's just no comparison with anyone else) and Japanese publishing industry is extremely strong, such that real experts can write somewhat technical books on semiconductor industry and you can expect it to sell if it's any good.
So nearly all of the "insider news" is going to be anonymous and read like rumor mill gossip. Especially in the English-speaking press, where there is an extra level of cultural and linguistic indirection.
That doesn't directly explain the poor quality of writing, but my guess is that the above factors dissuade a lot of talented journalists from even attempting to cover it.
(this is all speculation, of course)
What a weird hit-piece filled with assertions that could never be reliably backed up by other than Samsung themselves conducting a massive internal investigation.
"Clients and employees of SemiAnalysis may hold positions in companies referenced in this article."
So somebody seeks to benefit financially from selling the perception that Samsung is in trouble but feels an online disclaimer can shield them litigious Samsung lawyers.
This won't end well.
And if you think that the U.S. is not using every tool in the box to win markets, including bad-mouthing foreign competitors and lying to win customers over, then you're just lying to yourself. Have you been living offline for the last 20 years? The Internet is a battlefield and no-one plays more dirty than the U.S.
They lost Qualcomm, Nvidia, and Cisco for the next generation. Are you disputing this fact?
They did not fully ramp 1Z or 1 Alpha dram nodes. Are you disputing this fact?
The data centre parts were fabbed on TSMC 7, the consumer on Samsung 8
The consumer parts hit their performance envelope but use about 30% more energy than the data center parts, for the same performance envelope.
Samsung have removed the Elpis controller from the 980 PRO SSD and replaced it with an unknown one.
Take a look here for what's changed on the 980 PRO: https://i.imgur.com/YJShyLR.jpg
Companies want to build product identities that can last for more than a year, but also want to be able to react to changes in the supply chain. Fixed BOM guarantees really aren't viable in this product segment, but you can often get such assurances in the enterprise, industrial or client OEM SSD markets.
Most of the time, swapping in a newer generation of NAND flash memory is a net improvement for a drive's performance and power efficiency, but when the newer generation increases the capacity per die there will almost always be a downside in some corner-case benchmark. Those swaps aren't worth worrying about, unless you're trying to abuse a consumer SSD for the kind of workloads enterprise SSDs are designed for.
Likewise, controller swaps are usually nothing to worry about—aside from some instances where performance downgrades stemming from NAND downgrades have been misattributed to a controller change, the most harmful examples in recent years have been when the amount of DRAM is reduced from the usual 1GB per 1TB ratio down to something more like 256MB per 1TB. The consequences of such a change are easy to demonstrate with synthetic benchmarks, but almost impossible to measure let alone notice for real-world usage patterns. Other times, a controller "downgrade" simply swaps in a cheaper controller that is still more than fast enough for the NAND to be the bottleneck.
It would be nice if we could universally get more detailed spec sheets and a guarantee of a new consumer-visible model number when the major components change, but the consumer SSD market has proven too price-sensitive and the technology and supply chain too dynamic for that to be a competitive business strategy. For the most part, you do get what you pay for, except that a few of the top brands also command an undeserved price premium.
My pet peeve with the whole debacle is that words and promises no longer matter especially when marketing ends up using words like "upto."
Give me a minimum, maximum and specs written in stone and I won't care if you change out components. As it is now, the entire thing has become a game of whack-a-mole, SSD reviews are loosing their relevance and given enough time (+current trajectory) we could easily end up in SSD market which is more like fast fashion than anything else.
A fixed product identity for something that materially changes would best be described as a form of fraud, and it is unfortunate that sort of deception is business as usual in too many places.
For the most part, these changes are not "material" in the sense of having a significant impact on the overall performance of the product or its suitability for the intended use cases.
It's mostly the tech enthusiast audience that cares about these changes. That audience is highly susceptible to fixating on "objective" criteria or benchmarks that aren't actually relevant to any of their real-world usage. If you exclude the differences that only show up in synthetic benchmarks specifically crafted to reveal subtle differences in SSD performance, the scope of this issue and the potential instances of fraud are vastly smaller.
This article lacks the evidence to warrant the boldness of its arguments. And after reading through the paragraphs of poorly based assertions, I see this at the bottom of the article. Why am I not surprised?
I am under no impression that my articles can move fucking Samsung's stock. That's hilarious you think I could profit off the market by writing this.
There's plenty of evidence though. Be an industry insider and you'd recognize it all.
I do know that my reports have moved smaller companies stocks by 20% in a single day, and have been verified true in the past. - https://semianalysis.com/short-report-nvidia-supplier-cut-ou...
If I thought I could move the stock, I'd make the position in the morning alongside my clients, and publish shortly after, like I did with the article I just linked.
> I do know that my reports have moved smaller companies stocks by 20% in a single day
seems like a weird contradiction. you really think a disclaimer can protect you here? Like many of us pointed out, you've made quite a bold, unsubstantiated claims with poorly cited evidence. You've admitted that your writing in the past have moved the targeted company's stock.
> If I thought I could move the stock, I'd make the position in the morning alongside my clients
I'm sure you could but writing a piece like this, without substantial evidence, I think could influence investors opinions and you've admitted that your writings in the past have moved stocks negatively.
> There's plenty of evidence though. Be an industry insider and you'd recognize it all.
so where are the evidence that support your claims?
The cultural issues being the cause of these issues is the substantial claim.
A big part of Samsung's success is because of their insanely powerful and effective legal department. There's just no way their attorneys aren't reading semianalysis.
[0] https://www.hofstede-insights.com/fi/product/compare-countri...
That feels… counterintuitive, if true.
• Samsung: conglomerate Chaebol that includes of 50% revenue from shipping and ship-building, and then insurance, finance, travel, consume appliances, and finally smart phones and semiconductors which are minority revenue source
https://en.wikipedia.org/wiki/Chaebol
https://www.slashgear.com/831286/10-rare-samsung-products-yo...
• TSMC: single product company - foundry only
• Apple: "single product" company (compared to Samsung)
There is the Asian cultural side of it even for TSMC so it has many of the same top-down management issues. But the key difference is focus.
Taiwanese businesses are pretty much never Chaebol/Kieretsu-like - it's far more atomized, independent and interdependent, and focused on a narrow(er) market. There's also more "bottom-up" in terms of origins and this remains to some extent in many of them
(I work for a Taiwanese firm that bought my most recent company - and we work with direct competitors when it makes sense market-wise - that pretty much never happens with between Chaebols/Kieretsus. We also have a pretty selective market niche).
Compare this to Japanese kieretsu companies such as NEC - which is similarly broad but you also don't think of them anymore when it comes to semiconductor and especially DRAM.
Chaebols and Kieretsus are very similar in that they are conglomerates so they are broad with many unrelated businesses and not very agile especially since decision making is top-down. A key difference is Kieretsus are more likely to be top-down BEFORE and AFTER bottom-up consensus cycle - though that's a predilection based on organizational structure.
https://en.wikipedia.org/wiki/Keiretsu
The example the article gives of Hynix proves that it's possible to be a Chaebol yet give enough free rein to business units to make good decisions. But it can work against as well.
Another factor with Samsung: their shipping business has taken major hits in the last 5-7 years. Which given that is 50% of their revenue, probably still diverts proper executive attentions away from semiconductor and cell phones resulting is fiat decisions not based on good data.
https://thediplomat.com/2016/06/south-koreas-shipbuilding-cr...