Hacker News Reader

Comment by rkagerer | Hacker News Reader

We divested networking back in the ‘90s, we divested PCs back in the 2000s, we divested semiconductors about five years ago

Is it just me or does anyone else feel over the decades they've been divesting some of the best (long term) building blocks? A company with vertically integrated silicon, compute, networking, cloud, AI, Enterprise etc. seems like it could have such an edge if only they had focused those engineering capabilities on consolidated, high-margin end products.

I see the other big players going the opposite direction. e.g. Google and Amazon are building their own silicon for an edge in Cloud and AI.

So-called SexyIBM is just another cloud company without a distinguishing barrier to entry. Sure, their growth will look good on paper for a few years, but when Cloud becomes commoditized (which I think is already happening), the capabilities which could have created the kind of real innovation that opens up whole new industries will have all been cleaved away.

View on HN

IBM has not been a “company” in decades.

It is a financial engineering project. When you see it through that lens all of this makes much more sense.

Exactly, IBM is a company run by finance, not technical people. They survive despite, not because of what they do.

can confirm, IBM (GTS at least) has more people dedicated to ensuring you project your future billable hours and will directly reduce your performance rating for not providing the projections every 2 weeks than if you miss client deliverables consistently.

You're absolutley right about GTS, but it's clear that the services offerings from IBM aren't what top level commenter here is praising.

good thing its much easier to project future billable hours compared to actually delivering value quickly in 2020 ^.^

Not to nitpick, but the current CEO has a PhD in Electrical Engineering and started working at IBM Research in 1990.

I'm sure lots of people at IBM have EE degrees throughout it's existence. That hasn't caused the company to behave any differently. When approached by IBM to help my company build out our cloud infrastructure, we had salient measured targets and the reps (and subsequent "technical people to answer our questions") who presented were woefully inept. I was the Director of Engineering, what did they think they were doing pitching me butterflies and rainbows? What IBM has done for decades, like all the other parasitic remnants of industry (Oracle, Yahoo, arguably Apple, etc), and no CEO is changing that culture.

They are indeed incompetent, since a large chunk of competent people won't go there any more. But what kind of financial voodoo is keeping them afloat? Is it another Sears in waiting?

An enron Enron goes down in 10 or so years, an not so enron Enron might last a good 20 or 30 years.

Present day IBM is not very different from GE. It is a financial company trying to survive from profits in areas it monopolizes. Without anything else to do, GE is going down hard and now is being investigated by the SEC for "accounting practices", i.e., financial manipulation to keep the company alive.

They're effective as an Oracle-like consulting company that buys third-party products, integrates them into their platform, and then cross-sells them hard.

They seem to have a good research division. I learnt a lot form their research on neural hardware during my PhD

like all the other parasitic remnants of industry (Oracle, Yahoo, arguably Apple, etc)

Would you mind sketching the case for Apple as an example?

I don't know what they meant, but I think the case is reasonably easy, if the accusation is "financial engineering." Apple's overall range of products have shrunk over the years. I'm old enough to remember when they had 20 different kinds of desktop computers, updated every year. They now have 1 desktop, updated every few years, one laptop, updated every few years.

There is a sense they've given up on everything but the iPhone. Arguably, Apple keeps producing Mac computers only because iPhone devs need to run XCode somewhere.

The iPhone was introduced in 2007. It was very innovative at the time, though now it has dozens of Android clones. Total sales of iPhones seems to have peaked -- increases in Apple's profits have come from raising prices and managing margins.

This last bit would be the heart of the case that Apple is now increasingly financial in nature and therefore parasitic. They don't seem to be rolling out much in the way of new products, and certainly nothing as "change the world" as the iPhone was in 2007.

I think it is well known that Apple does not have the "division per product group" structure that almost all other large companies have. But Apple has an unusually narrow range of products for such a large company.

Even projects that Apple seemed excited about 6 or 7 years ago, such as home entertainment, no longer seems to be priority for them. I can tell you from personal experience, their Apple TV has been absolutely stagnant for years, and many of the problems, which I assumed they would eventually fix, have not been fixed.

Hmm. At first I disagreed with you, but when I think about it over time, I think I'm forced to agree.

Their newer MacBooks are going to be ARM based, which, I think, is mostly so that they have more people to develop apps for their phones and iPads. The rest of their customers will be buying them simply for day-to-day stuff wrapped in a pretty package.

I'm trying to think of a major Android phone manufacturer that does the same as Apple, but am coming up blank. Samsung, Sony, Oppo, etc, are all companies that have a multitude of products.

Maybe the only other comparison would be Google, but they were primarily a software company that is now making Pixel devices.So, not quite the same thing.

> Their newer MacBooks are going to be ARM based, which, I think, is mostly so that they have more people to develop apps for their phones and iPads.

I think it's oversimplifying the situation to say "Apple wants ARM just for mobile", though. While that's certainly a benefit of having a single architecture to support in their product ecosystem, there have been plenty of threads on HN about how Intel's rate of chip improvements have been stagnating. The much-touted scaleup of their 10-nm chips ended up being delayed by 2 years, and their 7-nm chips aren't arriving until 2023, by their own (likely optimistic) estimate[1]. The more likely narrative is that relying on Intel became untenable both cost- and performance-wise.

1: https://www.theverge.com/circuitbreaker/2020/7/23/21336356/i...

You are missing the iPad and their smartwatch. It is very important for a premium company to have lesser variety, but higher quality. I cannot respond to quality of their Macs, however the tablets, watches and phones are pretty well made compared to their counterparts. The only exception was old Nokia windows phones, but they do not exist anymore.

> They now have 1 desktop, updated every few years, one laptop, updated every few years.

Not true. They have three desktop form factors — Mac Mini, iMac and Mac Pro — and three laptop form factors — MacBook Air, 13" MacBook Pro and 16" MacBook pro. And each can be variously customised by CPU, memory, storage, etc. Most of these are updated with 1-2 year cadence.

Of course, it's much easier to make any case if you disregard the most basic facts within the first paragraph.

Um, sure. Those "form factors" are very different...

https://www.apple.com/mac/compare/?afid=p238%7CsboPJ4L5G-dc_...

> it's much easier to make any case if you disregard the most basic facts within the first paragraph.

Using apple marketing as a source of truth does not justify your snark. These "products" are small variations of grift.

I think that a case for having a very lean product line-up can be made from an engineering perspective as well. Having few SKUs and changing them rarely allows to focus on quality. It allows for very efficient manufacturing, since production is all about economies of scale. For already established product types, it enables iterative refinement. But perhaps most importantly it allows to put innovation focus on new kinds of products. The PC, laptop and tablet have all been invented and standardized.

With the AirPods, Apple has again managed to establish themselves as a leading player (in affluent market/segments). They did the same with smartwatches a while back. Their track record is strong. Probably will do the same with... $FutureProductCategory. Of course this is only partly due to engineering and product design - brand building and marketing also vital components. And I think it is fair to say, that while a strong engineering organization, Apple is not engineering-driven. They consider engineering as a means to deliver good product, not an end in itself.

I feel it has been a shift of focus from building new and innovative products to building supremely profitable products. It doesn't sound that big of a difference but in terms of how it affects the end user it is a noticeable change. You feel more like cattle that is being herded to buy this and that, very expensive stuff, knowing that something is off and it shouldn't be this expensive.

When trying to surpass their last year's profit margins they shouldn't do it mainly by increasing the prices. Even though the frog doesn't leap when you increase the temperature slowly, shouldn't mean you should do it.

Also look at Braeburn Capital.

It is a hedge fund owned by apple, it has at least $237B under management.

My guess is that he refers to the pricing/app store policies, and to the decreasing product quality.

The CEO is one piece of the administrative puzzle for a company that large (>100k employees).

IBM business units are black (Profitable or losing a small amount of money but missing targets), green (exceeding targets), or red (rapidly dropping revenue and losing money) blocks.

They buy a business, it's black. They don't understand how the block works, but they need to to be more profitable, so they paint the IBM logo on everything, cut costs and headcount until it is green for a short while. Then it turns black, they rinse and repeat. They make it green a few more times and then finally instead of green, it turns red!

Then they combine some red blocks, bundle them together, make it look black or green and sell them off.

This looks like the biggest bundle jettison ever, but after acquiring Red Hat, that makes sense.

Can you elaborate on that? I worked for IBM for a little more than a year and it was deeply disturbing, but I don't really understand what's happening under the hood.

Can you elaborate on what was deeply disturbing about your experience? Your comment has me curious.

They bought the company I was working for. They turned everything upside down. Tried to force us to sign contrats that would make them the owner of our public contributions which lead to an outrage so they backed down. Later they did "patent minig" sessions where they tried to extract good ideas from people for free. They also shut down all projects and pushed their Watson shit on us. They didn't give a shit about company culture, and people in general so 90% of the employees who had even a tiny bit of ambition left for a startup that was created by people who left IBM after the acquisition.

That pretty much sums up my experience with IBM when they bought our company as well. I had a colleague who quit immediately finding out they bought us out because he had worked for them before. I gave them the benefit of the doubt but he was right and I left once they started to integrate us into the borg. Just about everyone ended up leaving not to long after either.

Don’t you have golden cuffs? Aren’t you heavily incentivized to stay, at risk of not getting the money you’ve worked on for years?

When I was hired the company wasn't offering options and I negotiated a fair market price for my work so I was able to "just" leave.

Large companies allocate some percentage of their budget to "modernization" or "IT improvements" IBM tries to grab as much of that budget as possible and then to spend a little money as possible internally delivering results. So they end up selling "Watson AI" or implement a "24/7 Cyber Security Operation Center". They don't give a shit how useful or long term successful those products or services are. They only care about how big the contract is and how high their margins are on the contracts.

Why is IBM's stock soaring? This is the sign of a dead company barely walking.

Furthermore, IBM has been in this state since the 90s, right? How are they still lumbering around? Why do companies fall for their sales pitch?

IBM stock is performing terribly compared to other technology companies from 15 years ago GOOG,AMZN,APPL.

They deliver a good story to other inefficient companies. Those companies can tell their board that the "new Watson AI will optimize their operations in Q4"

IBM's stock is down 30% since 2013 and is only up 10% from its post-dotcom crash 2000 number. They pay decent dividends so it's been an okay place to park some money, but it's the opposite of soaring.

There is a ton of inefficiencies in big businesses and governments and IBM and other service companies are preying on that. It's a viable business model, if a little scummy.

That's a real shame. Most large, old American companies have fallen into that trap. Notably RCA (gone) and GE (going).

I hope Apple doesn't eventually succumb. Tim Cook's successor matters.

Apparently GE sold off trains which seems like a pretty good business to own?

Out of all the business lines GE has sold off over the years (computers, appliances, financial services, lighthing, television, etc.), why trains?

GE was a slow-mo train wreck for years (pun intended).

Recently read "Lights out", about their problems in the last ~20 years, fascinating read, heartily recommended.

And yet Jack Welch somehow came to be looked upon as a management guru instead of someone who gutted the long-term viability of GE in favor of financialization.

Luckily I think history is doing its job in changing how Jack Welch's legacy will be remembered. I don't think students or upcoming businesspeople are really learning Welch's approach anymore.

Jack Welch has definitely fallen out of favor considering the results obviously illustrate his failures.

They needed cash and it’s a profitable business so they sold it.

The business lines you offered as examples are low margin commodities or stuff GE really sucked at. Trains seems like a reasonable high margin line of business for a company that is actually good at industrial engineering.

If it was train lines including rail ownership I can see why they would want to drop out. Logistically that's a nightmare unless it's totally private.

Then again around here the rail lines are mostly owned by grain/grass farmer groups who ship product by rail and can load from their farm directly. Passenger lines that share the rails wait for those trains.

It was the business that builds trains.

GE was a big player in the diesel-electric locomotive market, starting with the GE Universal series, then the Dash 7, Dash 8, Dash 9, and Evolution. The division was sold off at some point there, I think before the Evolution. Back in the earliest days the engines themselves were provided by companies such as Cummins but in the 60s GE started building its own prime movers.

As stated by another post, they built the trains but didn't own the track.

However, I'm not sure your conclusion about owning railroads is a nightmare is true. Berkshire Hathaway (Warren Buffet's company) bought Norfolk Southern in 2007 and it's done fairly well since. My understanding is that these are relatively stable businesses because the barrier to entry is large enough to stifle competition.

GE Transportation still exists, just in a different form. It's called Wabtec: https://en.wikipedia.org/wiki/Wabtec_Corporation

Just because it didn't retain the name "GE" doesn't mean the business is gone.

Yes it exists but GE doesn’t own it.

but is it high-margin, high-growth? The same thing is happening in Germany with Siemens for years... - just 10-20 years later than the US. Once they did everything (for real): car parts, control software, nuclear reactors, trains, solar cells, semiconductors, computers (even mainframes), medicine products, household appliances. Nowadays: BLOCKCHAIN. I guess the job of a chinese "we will rule the world"-planner is a lot easier through western greed, than it should be.

>but is it high-margin, high-growth?

They sold trains and bought into oil & gas exploration. If high growth is what they're aiming for I'm not sure they picked the right horse.

They still do almost all of those things. Look at their yearly report. They are also making huge profits from them.

Apple is already on it's way to the trashheap. There's a reason why innovation has essentially stopped and now comes down to "a few megapixels more in the camera", while quality control - both in their hardware as well as their software - has taken a big hit.

I guess they will succumb faster than the old Xerox-age market leaders, because they are not focussed on consulting (= making other companies believe their constantly syphoning money from them brings value).

>There's a reason why innovation has essentially stopped and now comes down to "a few megapixels more in the camera"

How is literally designing your chips to a point where they could be desktop class to replace x86 not innovation? What other company is doing this?

I can't imagine a "finance company on the out" deciding that they will bring all chip development in house.

I recently started developing my own chips and put them on all my desktops. Thinly sliced potato, hot oil, a bit of salt… Delicious.

Seriously though, some people just don't like Apple and they make a lot of noise about it.

> How is literally designing your chips to a point where they could be desktop class to replace x86 not innovation?

That's more a sign that the company wants more profit by owning the top to bottom stack.

Perhaps they saw the existing chipsets as not delivering what they wanted or not scaling to fit demand, but it's still an investment not directly tied to product (their core competency).

> That's more a sign that the company wants more profit by owning the top to bottom stack.

They pretty much had their hands tied. Intel failed to deliver for years now, and AMD never had a competitive offer on mobile and still does not.

> That's more a sign that the company wants more profit by owning the top to bottom stack.

Is bringing mobile/embedded and now desktop-class CPU design in-house really something one does as a cost-saving measure? Apple wants control over their entire stack, and sure, that relates to their business as a whole, but if this was solely about profit maximization surely there would be better strategies.

> it's still an investment not directly tied to product

I'm not sure I follow your reasoning here. Are you arguing it's not a direct investment because the CPUs aren't products in and of themselves, but rather components for other products? If so, I don't agree -- Apple's investment in, say, case tooling/manufacturing processes and equipment exclusive to their products is surely an investment directly tied to those products, right? The CPUs are likewise components exclusive to Apple products. That seems to me to be a pretty direct investment.

I don't understand this response at all. Does innovation not count if you're not doing it for charity? For several years I was reading articles about how Moore's Law was totally over and we couldn't expect any more improvements in chips, and then along comes Apple to blow x86 out of the water.

> but it's still an investment not directly tied to product (their core competency)

I don't even agree with this- Apple's core competency is the top-to-bottom customer experience, which they (almost certainly correctly) think they can improve by making their own silicon. But even if it was true, so what? Again, "investment not directly tied to product" doesn't make innovation "not count".

> an investment not directly tied to product (their core competency).

Heavily disagree, the only reason I would renew my mac in the next 6 month is because the ARM chip, and I will as soon as it comes out.

me too

>How is literally designing your chips to a point where they could be desktop class to replace x86 not innovation? What other company is doing this?

IBM for one; https://en.wikipedia.org/wiki/Z/Architecture

I mean you asked, and it was obvious. Z is a more interesting architecture by far than the turd Apple is shipping.

>IBM for one; https://en.wikipedia.org/wiki/Z/Architecture

IBM has been building chips since day one. I'd expect IBM to divest their mainframe business eventually. The distinction you are missing is that usually "finance-driven" companies don't usually decide to pour billions of dollars into bringing in an already outsourced component in house that they hardly have experience in.

Furthermore, I don't see IBM using the Z to "innovate" - They aren't pushing the mainframes to anyone other than people who are already buying mainframes.

>Z is a more interesting architecture by far than the turd Apple is shipping.

The Z, an architecture for people who are pretty much already buying mainframes, is more interesting than a desktop class chip with what will probably be a completely unmatched in performance/watt? I don't see how the Z is more interesting than a chip that is finally attempting to challenge the 30 year x86 dominance in desktop computing.

Both of IBM's current chip architectures, Z and POWER9, are indeed interesting.

Apple hasn't shipped their "desktop class" architecture yet.

Perhaps you are frustrated by the design constraints of low power mobile chips. Ok.

But if you're paying attention, architecture wise, it may be of interest to note that Apple's ARM chips, so far, have delivered good performance in their handheld applications by careful attention to sustained memory bandwidth. Competitors went with more CPU cores.

So there's some fun chip architecture to be had, even in 2020.

A desktop Apple architecture might use something like HBM for main memory, rather than DIMMs.

There's lots of room to innovate, out there in consumer computing.

Memory bandwidth is generally kind of needed to take advantage of more cores.

Last time I checked, Z had 500gb/sec; more than 10x Apple's. Kind of wish IBM had won processor wars. Generally speaking whenever I look at their mainframe doodads, then look at the hot garbage being slung over at Amazon or whatever FAANG shit hole, it makes me sad. The company with the best engineers is an also-ran that mostly sells consultant hours. Maybe they'll sell off the mainframe business independent of the rest of the horse shit and it will undergo a renaissance. Doubt it though.

> How is literally designing your chips to a point where they could be desktop class to replace x86 not innovation? What other company is doing this?

Facebook, Google and Amazon are known to do their own server design; it wouldn't surprise me if any or all of them were doing custom processors (e.g. better virtualization features for their clouds, or processors that are more oriented towards their workloads). It doesn't sound like innovation, more like cost cutting; these processors aren't delivering a step change to end users, at best they're squeezing out a little more battery life. (By contrast e.g. that sapphire screen that was rumoured would have been innovative, because sapphire can do stuff that glass simply can't).

“Wouldn’t surprise me” is a damn long distance away from “has put actual dev kits in the hands of developers”.

Those companies don't need third parties to develop anything so why would we know? Facebook was building custom server hardware for years before it became public knowledge that they were doing it.

You're right — we don't know. That gives us two options. Either we acknowledge the innovations we do know about and have proof of, or we use wild guesses and assumptions to dismiss those innovations.

I'll be happy to praise Google or Facebook for advancements in CPU tech if and when they show us such a thing. Until then, publicly available facts are that Apple is innovating in that field and they're not.

Actually we do know that Google and AWS are investing heavily into custom chips.

Actually we do know that Google and Amazon are investing heavily into custom chips.

The very fact that we can't tell shows that this isn't any significant innovation. Even the part about handing out dev kits doesn't actually show anything - using an off-the-shelf ARM would create the same impact.

They're designing CPUs - something that many companies have done and many companies will do. Big whoop.

There's 2 trillion USD betting that you're wrong, and virtually zero betting that you're right (AAPL Short Percent of Float = 0.00% as per Nasdaq). Buy some puts and you'll make a killing.

Not sure why this is being downvoted, Apple’s hardware reliability has become horrible.

Every Apple device I’ve purchased in the past few years has suffered from a defect: AirPods Pro, iPhone X, 2019 MacBook Pro, iPad. And this doesn’t include Batterygate.

Why's it being downvoted?

> Apple is already on it's way to the trashheap.

That's why. Even if they're lost some of their lustre (and, if we're being literal, could be "on it's way" in the sense that it was the most valuable company in the world and might've dropped a few percentage points).

Not only the past few years, heck, they launched a phone you couldn't hold properly and people lapped it up. That's not stopped their offerings not only being the best in their class, but often the only products of note (AirPods, iPads, Watch...).

Apple doesn't have to beat Apple. They just have to beat the best of the rest - and apart from perhaps Samsung and Huawei in phones, they're looking pretty peachy still.

Contrary to popular knowledge, free market companies don't grow to take over the world. They grow until they get strangled by their own bureaucracy, and then rot away to be replaced by a younger, nimbler company without the bureaucracy.

History is full of examples. Just look at the top 10 American companies by market cap, decade by decade.

As opposed to non-free market companies, that stay even when they are strangled by their own bureaucracy.

Non-free market companies are propped up by force.

Sounds like an interesting exercise. Let's look at the ten largest American companies by market capitalization in 1960.

1. American Telephone & Telegraph is now called AT&T and was in the top 10 within the last decade. 2. General Motors has dropped off the top 100 for a while, but was the world's largest automaker within the last decade. 3. E.I. du Pont Nemours dropped off the top 100 in the last few years, but was the world's largest chemical company within the last decade before selling off Dow. 4. Standard Oil Co of NJ is essentially ExxonMobil and has been the #1 largest company within the last decade. 5. General Electric has been in the top 10 within the last decade. 6. IBM has been in the top 10 within the last decade. 7. Texas Company is now called Texaco and is part of Chevron, which has been in the top 10 within the last decade. 8. Union Carbide had one of the world's largest industrial disasters, and its later history is involved with duPont above as part of Dow. 9. Eastman Kodak ceased being relevant in the last 20 years. 10. Sears Roebuck & Company has collapsed, last being the largest retailer in the 1980s.

tl;dr: Of the top 10 in 1960, half have been in the top 10 within the last decade (including the former #1) and one has been #1 within the last decade.

Keep going back in time. RCA, for example, used to be the biggest market cap company. Now, few people even know it ever existed.

I do plan to go back farther in time; I think it will also be interesting.

The claim about RCA seems unlikely. Radio Corporation of America stock grew very fast through the 1920s, peaking in September 1929 before the crash. However, even at the very end, it doesn't seem to even have been top 10 among industrials in the S&P index: http://piketty.pse.ens.fr/files/McGrattanPrescott2001.pdf#pa...

Instead, on that list we have companies like General Motors, General Electric, and Standard Oil of New Jersey that were still big in 1960 (and as mentioned above, some still so in the last decade).

How much of this is due to changes in the nature of the economy vs mismanagement? In the 50s and 60s the largest companies were focused in manufacturing and oil. Now the largest are in tech because the economy is different.

GE is an interesting example because, as a conglomerate, I would think it's better poised to strategically change the focus of it's business...maybe that's the point of the IBM spin-off.

Yes. They aren’t in the innovation business. They are in the Marketing (Watson), audit and share buyback business.

IBM used to be an engineering-first company. But people are expensive and engineers are some of the most expensive people there are. So when the bean counters took over IBM, they decided to get rid of their expensive people and replace them with cheap, fungible labor. Then IBM embarked on a series of experiments about how they could continue to make money with said cheap labor. This is the result.

Very well put. The same management philosophy happened to Boeing, and I think more quietly to many companies thought too big to fail in the last 50 years.

The difference on impact is interesting. Boeing is an important company because they are one of a tiny number of companies that makes the world's commercial airplanes. Arguably their transformation into a useless junk-heap of a company that can't build good products has been underway for a while, and that's a huge problem for aviation.

IBM is one of many, and their decline wouldn't really matter all that much in the grand scheme of things.

Since IBM bought RedHat, the parallels between IBM and Boeing here are too close to comfort. RedHat is potentially as vital to the Linux ecosystem as Boeing is to the (pre-pandemic) civil aviation sector. If IBM falls apart and RedHat vanishes, the impact on the Linux sector could be almost as bad as Boeing collapsing. Since Linux runs everything that's not a desktop computer or an iOS device, IBM-RedHat failing would be a very major problem for everyone.

I think Suse and Canonical would pick up the slack.

Explain how RH could vanish even if IBM fails/falls apart. I don’t see any scenario where that would happen, sorry. I’m not sure you understand just how entrenched RHEL is in ISP/telecom/Enterprise/finance/banking/insert-pretty-much-anything-here is..

RedHat revenues likely aren't enough to keep all of half-IBM afloat. It's vaguely possible that IBM could go into Chapter 7 if things go badly enough. That's the worst-case scenario I was speaking to.

Ok I get that part, but surely Red Hat/RHEL itself wouldn’t perish as a result?

If a company goes into chapter 7, the entire company is disbanded and its assets are sold off. In that scenario, Red Hat, as a division of IBM, would cease to exist.

Perhaps. My point is that Red Hat will still thrive under whatever parent company they wind up in. Red Hat would still be a thing. A very huge thing, and still a very huge moneymaker. Red Hat doesn’t disappear.

Linux will be fine, people who used RHEL can just moved to cloud and/or Amazon linux or Canonical, whatever that suits them.

The changes could even be by the same people, I'm sure many from RH will not like IBM culture.

RedHat makes much more then just a Linux Distribution, for my part i fear for Ceph.

This is one of the huge selling points for RH IMHO: Since they are open source if the vendor dies the product doesn't. Ceph is an amazing product with some really big customers, it wouldn't go down with the ship.

That said, Red Hat is doing well and I don't see them dying.

Disclaimer: I work for RH but opinions are my own

Last I heard Amazon itself was struggling with migrating from RHEL to Amazon Linux.

Red Hat heavily supports OSS projects used on their distributions. No other Linux companies contributes so much. (SuSE would be second, Hey Amazon?)

The pitch must have been like "hey guys, do you want to make some products, or do you want to make some money?"

Hard to create value when ditch or reduce power from the value creators and end up nothing but value extractors.

Companies taken over and run by value extractors (finance/business/marketing) over the value creators (engineers/product/creatives) end up in this stagnated, picked apart state when the grace from the value creation or product wears off.

Value extractors need to learn that you must first create value before you extract it. The reason value extractors originally were attracted to the project/product is usually that it has created value.

R&D has very little value to an MBA so it is cut, but long term it is all the value of the company. Value extractors kill the whales before they even can grow up.

> IBM used to be an engineering-first company.

I would argue that this was never true--even from inception. HP was an engineering-first company--not IBM.

IBM was about sales and marketing--they would rent and finance equipment for you even way back.

Now, IBM had world-class engineering, but people forget that a lot of the major companies had great engineering until the 1980s.

This is very true. It was even proudly stated in new employee orientation when I joined the company in the late 90's.

It was always sales oriented and engineering was something they had to do to deliver some of the things they sold. This may be the first CEO that was actually an engineer.

Thomas J Watson Sr. was a salesman.

I think they have enough money that they no longer have to take a risk in research and development. Their strategy is to wait to see what companies are coming up or have a good position of a sector , then they buy them out. Less risk , less overhead, and lower salaries for them to not innovate internally . Plus they get a new client portfolio from the acquisition and new talent.

Having worked for IBM and being a client of IBM in the past with large gov't agencies , i can also tell you that part of their success is liability perspective from the client. I know many of gov't big wigs who simply hired IBM not because of their talent but because if TSHTF they covered their arses with congressional hearing or lawsuits by saying what more could we do than hiring the vendor that made the product.

Nobody ever got fired for buying IBM only works until someone does.

And that turning point is happening/happened.

> Is it just me or does anyone else feel over the decades they've been divesting some of the best (long term) building blocks?

IBM is old enough to understand the pitfalls of that approach.

In year zero it sounds great. Integration! Efficiency! And that works for a bit.

Then internal politics causes you to pass on opportunities that competitors take because taking them yourself would cannibalize sales in one of your other lines of business.

Then somebody else comes up with a better microprocessor than you, or better software, or lower cost manufacturing, and it's going to take your internal team five years to catch up. In the meantime all your systems are integrated with your in-house solution, so you can't switch, but the deficient in-house alternative makes you less competitive, so you lose all the customers who can switch easily.

Then you have to cut R&D because of the lost revenue and fall even further behind the competition, which leaves you in a death spiral where the most profitable strategy is to lock-in your existing customers and milk them as hard as possible as you circle the drain.

Or you can concentrate on doing one thing and doing it well, so that you don't risk sinking your entire operation as soon as you fail to execute flawlessly in any one of the six different major fields you've created an internal dependency on.

Yeah, successfully vertically integrated companies like Apple choose very carefully which things they integrate vertically and which things they don't. Apple silicon for example isn't truly apple silicon. They still cooperate with a company for the manufacturing. Apple still buys gorilla glass from Corning, etc.

And in the areas where they do compete, they pour in top dollar to make actually competitive products.

It's way too early to tell how Apple is going to end. They've only been where they are for a small number of years. What happens if Zen3 turns out to be significantly faster than Apple Silicon, or next year Intel gets 7nm right? Do they abandon it or do they fall behind? What do they do if Google gets worried about their market share and decides to start selling a top of the line phone for cost and marketing it on google.com?

And the opposite failure mode is just as bad -- you actually succeed in monopolizing the market and then get smashed by antitrust.

Apple has dominated high margin consumer electronics for at least 15 years now. iPod, Macbook, iPhone, iPad, Watch, Airpods...

No one has the budget to compete

>And in the areas where they do compete, they pour in top dollar to make actually competitive products.

Apple TV+, Apple Music, Apple Arcade, Apple News+. Every single one of them were done to mask the profits of their Services Revenue. ( You could argue they provide decent value if you think they are competitive. )

I think that is where it all start to feel like Apple wasn't actually going in because they want to do better. They are doing it for numbers and finance.

And it irritates the heck out of me those money aren't being put or spend on better product and services ( on the actual product ).

IBM still designs their own silicon.

They still make chips in the POWER architecture and the Z architecture for mainframes.

https://en.wikipedia.org/wiki/POWER9

https://en.wikipedia.org/wiki/POWER10

https://en.wikipedia.org/wiki/IBM_z15_(microprocessor)

Additionally, the article doesn't mention IBM research, which I assume will continue to stay with IBM.

IBM is a very capable company when it comes to technology R&D. But they've always seemed to struggle to develop those capabilities outside of high-end consulting work.

IBM had the technical chops to create the likes of ARM, nVidia, or any number of big tech companies. They just never seemed to have the leadership capabilities to get out of their comfort zone. I understand the idea of keeping a business focused, but Alphabet and MS don't seem to struggle with managing a diverse portfolio of companies.

I wonder how IBM looked at 22 and 45 years old.

At 45 I guess they were expanding into Europe.

Yep, selling Hollerith card equipment to Germany in the 30's and 40's.

That math doesn’t work out. IBM was formed in 1911. 45 years old would be 1956.

Also doing business in Europe in the 1930s doesn’t mean that’s when they expanded. That business could have been decades old by then.

Some quick reading of the history of IBM and subsidiaries suggests they were doing business in Europe by the 1920s.

The answer probably depends on where you define the "birth" of IBM - the founding of the companies that merged to create IBM, the date of the merger, or the date they changed the name and/or began producing business machines with the name "IBM" on them.

In the context of the thread, it's reasonable to say that both at 22 and 45 IBM had strong leadership (Watson Sr.) that successfully managed a diverse group of subsidiaries. Although in neither case are there any products like a modern digital computer involved.

Like Nazi allies, according to some Holocaust survivors: https://en.wikipedia.org/wiki/IBM_and_the_Holocaust https://en.wikipedia.org/wiki/IBM_during_World_War_II.

It's not like the Holocaust survivors were the ones in the the commuting offices witnessing the deals.

Maybe in 20 years we talk about Apple/Microsoft/AMD etc like that...but it's not Germany but China.

IBM failed to leverage their silicon designs for driving down their cloud costs. As long as the IBM cloud runs mostly on third-party hardware they have no sustainable competitive advantage.

Sony PS/3 was a bold move though in that vain.

Running on third-party hardware is leveraging comparative advantage. Like Itanium, POWER is more expensive and there are very few reasons you would want to run on it. Definitely not a good fit for Cloud.

Full vertical integration isn't necessarily good business. Even Apple doesn't do its own manufacturing.

POWER is more expensive for us to buy, but IBM pays for it by the square inch of silicon. By that metric, it's probably very competitive with other server platforms. With OMI, they can also offer larger instances than anyone else while having more flexible tenant allocation because a single node can host a lot of tenants of different sizes up to the size of the full node, which could dwarf anything AWS and Azure can offer.

They offer their POWER machines as AIX and IBM i environments, as well as IBM z as LinuxONE VMs (under KVM instead of z/VM) with encryption everywhere for the security paranoid.

I suppose a z15 LinuxONE cloud offering would be attractive for them to offer because of all the built in observability, and also because those cores can be ridiculously oversubscribed before anyone notices.

As for pricing, they wouldn't need to pay their margins unless they eat into their other segments.

I've been trying to figure out whether the mainframe business is being spun off or retained as part of this split. Haven't been able to figure out whether IBM even manufactures this equipment. I know they outsource the chip fab now. I'm not even sure what the mainframe is called any more, they seem to change the name every couple of years. IBM Z? zSeries? System Z? Something with a "Z" in it I think.

Their latest is the z15 and their branding refers to IBM z (the same way they call their POWER lines IBM i, for the descendants of the AS/400, and IBM p, for the POWER boxes that run PowerVM, Linux and AIX). They build them mostly to order, in-house.

IBM is just a conglomerate of neglected acquisitions that share branding ("watson" etc.). It hasn't been a single coherent company in decades.

> We divested networking back in the ‘90s, we divested PCs back in the 2000s, we divested semiconductors about five years ago

Motorola followed that path. They sold off their computer business in the early 1990's. Analog electronics (On Semi), and then the semiconductor business (Freescale, now NXP). Finally, they split the government biz (Motorola Solutions) from the phones, and that was first sold to Google, and what remained then went to Lenovo.

Motorola was a gigantic powerhouse, and through mismanagement, is a shadow of what it was. Such a shame.

It's funny how back in 2005, IBM sells their PC/server division to Lenovo. Lenovo is currently the largest PC maker. During this time Dell has been bought and sold and re-bought.

Lenovo is chinese, right? How many other (big) chinese companies have been through Dell's cycle as you describe it?

Where is Dell making its computers? When I track shipments they seem to be coming from China.

I don't think they make their own hardware. They design it and then farm out manufacture.

What do they even design? All of the components inside their computers are off the shelf products from other companies.

Speaking of their servers, their motherboards and chassis seem to be of their own design. Often times these layouts are pretty proprietary and designed exactly around the design of the chassis. They do some level of customization of firmware on the devices they go with, as there's usually Dell firmware on things like hard drives which integrates more with their own management tools. Arguably their software stack of actually managing fleets of their hardware is something they design, along with the actual remote access controller systems they put on the boards.

Dell desktop workstations have custom cases and mobos and power supplies, at the very least. Laptops are custom cases and often SoC.

Clayton Christensen has written and talked about how Dell sold their business piece-by-piece. Each individual step seemed like a good move - it lowered costs and increased profits. Until one day the companies they outsourced to had it all and they started selling better computers for less money than Dell.

https://www.forbes.com/sites/stevedenning/2011/11/18/clayton...

It's interesting that the difference between that and Apple is that since Apple controls the OS, nobody else can make a Mac.

Unfortunately for Dell, anyone can make a PC.

Lenovo's margins are still razor-thin. The thing that bothers me most about IBM keeping POWER and Z is that I can't afford either, because the margins are eye-watering.

Fine, thin margins. So grow your market share and revenue, which is exactly what they did.

Since ~2005 they have ~5x'ed their revenue. Which is pretty significant if you consider one of IBM's big problems has been declining revenue. Proving the problem with the business wasn't the business itself but IBM management.

"Largest PC maker" has been a pyrrhic title for many years now. Lenovo's market cap is $8 billion, Dell's is $51 billion.

Yep.

And the value of Dell's business isn't in selling PCs, it's in other services around the PC.

And IBM's is currently $117 billion.

Most of Dells market cap is the 87% of vmware it owns (67 billion)

>I see the other big players going the opposite direction. e.g. Google and Amazon are building their own silicon for an edge in Cloud and AI.

IBM still designs chips - what they divested was their semiconductor manufacturing (fab) business.

They actually design tech that goes in other companies' chips.

That was my first impression too. It was extraordinary that that comment was said positively by the CFO, rather than critically by a shareholder or journalist.

Apple, Google, Amazon are all proving the value of vertically integrated product lines, while IBM chooses short-termism.

They really had no choice about divesting PCs. IBM was on Microsoft's bad side so was being charged more for Windows than competitors, and PCs already had razor thin margins. After they sold off their PC business, Microsoft had no reason to charge extra for Windows and that became a viable company.

> IBM was on Microsoft's bad side so was being charged more for Windows than competitors

I thought the DoJ investigation stopped Microsoft from doing such things. What am I missing?

It was a slap on the wrist.

https://www.justice.gov/atr/case-document/file/503541/downlo... has the judgement. Read the following phrase and consider the size of the truck you can drive through it.

Nothing in this judgement shall prevent Microsoft from providing Consideration to any OEM with respect to any Microsoft product or service where that Consideration is commensurate with the absolute level or amount of that OEM's development, distribution, promotion, or licensing of that Microsoft product or service.

When IBM was competing with companies that got such "Consideration" and it couldn't, guess how that worked out?

IBM killed their own PC business by going with the MCA bus, and not allowing 3rd parties to make expansion boards without a license. And the PS/2 series was quite expensive at the time.

Vertical integration is high risk, high reward in that if each component in the stack is a market leader in their industry, the overall company becomes exponentially more valuable.

On the other hand if any component is non-competitive, the other components in the stack are still obligated to give that unperformant component their business. Which reduces the pressure to perform and causes the overall company to gradually become less competitive.

I think a good example of this would be Intel designing chips and also owning the fabrication plants. It's no secret that Intel has been to fabricate their chips for a while now.

Like Boeing there are too many MBAs and not enough engineers in charge!

Yeah! Why differentiate in quality, when we can blame all holders of a type of degree?!

It's perfectly fair. If they had hired all chefs into upper management, one could say "too many chefs and not enough engineers!", regardless of the quality of the chefs.

Only if you think "Master of Business Administration" and "Chef" convey equal amounts of business administration skill. The exective board page linked above is 22 people, even if all of them held MBAs could it really be that 22 expert business administrators is too many for a global company with 350,000 employees?

It's not about skills, it's about priorities.

Their job is to make money; they don't care, and they are not paid to care about anything else.

Now... the question is, does someone with a different background care about the company vs making money in different proportions to the set of 'expert business administrators'?

I posit that yes, someone with a different background will have different priorities to simply 'make as much money as possible right now'.

Maybe even a chef.

As a % top leadership, yeah it totally can be too much.

"The secret of Costco’s success revealed! (hint: no MBAs need apply)", https://washingtonmonthly.com/2013/06/09/the-secret-of-costc...

"... Costco does not hire business school graduates—thanks to another idiosyncrasy meant to preserve its distinct company culture. It cultivates employees who work the floor in its warehouses and sponsors them through graduate school." https://www.bloomberg.com/news/articles/2013-06-06/costco-ce...

I see this type of comment thrown around a bit, but a quick pass of the senior management suggests otherwise. https://newsroom.ibm.com/executive-bios? Not a whole lot of MBAs and a decent amount of engineers.

Ask those managers when the last time they wrote a line of production code was :)

Even the most technical CEOs like Bill Gates stopped writing code pretty early in their careers

When's the last time Tim Cook or Satya Nadella wrote a line of code?

Something about cherrypicking outliers compared to the droves of managers who haven't written code in 20 years and have been converted to the dogma of management

If you loosen your sense of "MBA"...

- Rometty (Chair): Technical degrees, held technical positions for first 10yrs at IBM, then in sales for the 90s, two decades working primarily with finance customers and she worked the PWC acquisition.

- Krishna (CEO): EE, Idea guy.

- Whitehurst (Pres): MBA

- Boville (SVP Cloud): Business degrees (UK)

- Browdy (SVP Legal): IP attorney

- Foster (SVP Services): Art degree, Accenture guy

- Gherson (SVP): Management degrees

- Gil (Dir IBM Research): EE/CS

- Kavanaugh (CFO): MBA

- Got bored

One trait is pretty dominant: decades at IBM

The loosening lens is a bit of strawman argument. I would hope someone at the level has experience with the business side of things... Pichai and Nadella weren't tapped to run their business straight after shipping a release.

Also, Howard joined in May (which is an important one...because this press release is about Cloud not about HR/patents.)

You don't need MBA to have MBA mindset. A lot of computer science grads use their degree to get foot in the door and then move into management positions.

At IBM, to do anything you need to get permissions from 5 different semi-tech approvers. They have not done any real work in a while but read a few blog posts and come up with their own policies that contradicts each other. It is a pita to want to produce high quality code.

And that is just middle management. All those people higher up can define tech words but hardly anyone can actually explain the definition of those tech terms.

Ironically Whitehurst came from Red Hat.

Whitehurst has a Computer Science degree (BS) and was using Linux on his personal computers in the 1990s and early 2000s, I think it's unfair to drop him in the MBA bucket.

Vertical integration is a hot term, but unless you have long term monopoly in the market, it’s more likely to be a losing strategy. If you bet on wrong direction, you’re in much deeper troubles, as you cannot switch to other vendor that went other direction (switching vendors is also hard, but much easier than rebuilding and catching up on your internal R&D).

And you lose power of market and economy of scale. Vendors can focus on developing their tech to be better, for everyone, as they get bigger R&D budgets, and you can just focus on improving things that are your true market advantage, not fighting in commodities market.

Note, that doesn’t mean that IBM is doing right thing. But vertical integration is extremely overhyped.

Yeah. Mainframes are arguably the most vertically integrated products there are and yet they’re a dying breed in a rent-maximizing at best industry than a growing one.

As a former IBMer, I completely agree with that sentiment. Little of what IBM is good at is sexy but that doesn't change the fact that they lead on many vertically integrated things. Turning back on those things will hurt them in the long run.

IBM still has plenty of sexy hardware stuff - mainframe and POWER (AIX not so much) are pretty impressive.

That z/OS has tens of millions of lines of code surprised me the other day. It's not like it needs to support 23 brands of mouse and dozens of GPU architectures.

It's backwards compatible to really old versions. It wouldn't surprise me if 70% of that code was emulators for previous versions of the OS.

Even IBM S/360 is still being emulated.

There’s not much compatibility for old versions of the OS - IIRC anything older than the hardware itself is not supported at all. Backwards compatibility is for user code.

IBM is going to fade away.

They're divesting lower performing (but not money losing) ventures in order to not hide the performance of their higher performing pieces. This is so they can show the unsustainable growth that stock holders (still) demand from corporations.

They'll look great on paper for a while as you say, at least until their cloud business fails for some reason. When you average growth over a whole company, you do mask high performing parts with low performing ones... the same is true of cash flow and revenue.

If you get rid of parts of your company that are solid performers but aren't doing as well now, then you're getting rid of some of your safety net in the hope you won't need it.

IBM's most impactful business (in terms of wide use) is probably their research labs. Many, many chips today use SOI and copper interconnect technologies. That sort of research doesn't apply much to the cloud, but maybe they're hoping their AI stuff will.

That sentence struck me as well, mostly because people continue to make tons of money in networking, "PCs" and semiconductors. It's not that they didn't "fit the integrated value proposition", it's that IBM has terrible leadership.

No bets needed for what band of clowns will lead the hip, "value proposition" part of the company of course.

How about this then: IBM realized they're a bunch of clowns and can't compete in these industries so they made the smart move and got out.

Vertically integrated monopolies are inefficient because practically, what you really have in a vertically integrated monopoly is a bunch of separate companies, all following a single, centralized entity and paying tribute to that entity. There is an opportunity for profit here, for shareholders, but very little room to actually improve technology, provide utility to users. Every product in that monopoly becomes less of a computer, more of a mouse-trap to keep people in the ecosystem. If your monopoly has even a single component which can interoperate with other monopolies, then you're basically subsidizing a whole layer of infrastructure.

Plus this practice paints a huge target on your back for anti-trust action, as you are seeing now with the other monopolies. The bigger your monopoly gets, the more potential profit there is for whatever government agent can bust you up. This could result in promotion, christmas bonus, maybe even a corner office with a window. Very high stakes.

Yes, this quote stuck out for me immediately as I thought of the custom network hardware run at Google, AWS, etc.

The pieces they've divested haven't been valuable as part of IBM. They've fit their new owners businesses better. Thus they've been more valuable with their new owners, which is why IBM sold them.

> Is it just me or does anyone else feel over the decades they've been divesting some of the best (long term) building blocks?

Staying at home, the best thing I found to play with is built "digital lego" MOCs using LeoCAD app + LDraw Parts Library.[0]

Seems like IBM CEOs, stayed at home for a long term, just found own "lego-like" game to joy ;)

[0] https://github.com/Symbian9/AWESOME-LDraw

> A company with vertically integrated...seems like it could have such an edge

Yes but I guess it was the IBM board whom decided that they could better invest in the money with various initiatives. IMHO, they are now (and for ~15 yrs) only choosing a safe route, not an innovative route.

The parts of IBM are worth more than the whole, and many of the company's board members and investors realize that. They are getting the value of these deals, even if "IBM" is not.

Couldn't agree more! I actually believe that only the divested business unit, NewCo, will survive

Let's call them Elppa Corp.

I prefer HAL Corp.

anecdotally, my nick is an application of the HAL shift to my previous company

Reply on news.ycombinator.com