Hacker News new | past | comments | ask | show | jobs | submit login
The technology industry is rife with bottlenecks (economist.com)
84 points by mariushn 4 days ago | hide | past | web | favorite | 40 comments





I'm still royally $EXPLETIVE_HERE about Analog Devices buying Linear Technology a couple of months ago. AD and LTC used to be the mutual second source suppliers for critical components in lots of my projects. For almost every part AD had there was a matching counterpart in LTC's portfolio and vice versa. And often enough I could design my PCBs in a way that they even could be populated with either chip, depending on availability. Most of the ICs are still available in their AD and LTC variants, but I wonder for how long.

I've similar feelings about the acquisition of National Semiconductor by Texas Instruments (primary and second source for high speed ADCs).


Ah, Analog Devices, the king of "expensive because we say so" parts.

If I had a lot of money (which I don't) I'd have started my very own analog semiconductor design and fab company years ago, with the intent aim to be a direct competitor in the market of high speed and precision analog and signal conversion electronics.


Thank you

The thing that truly worries me is that once these countries are no longer dependent on each other it will be easier for them to get into war with each other.

At one point people claimed "no two countries with a McDonalds had ever been at war with each other"; this was I believe ended by the NATO attacks on Serbia in support of Kosovan forces.

It seems possible that the US will get itself into a conflict, possibly (hopefully!) non-total, that disrupts its globalised supply chains and results in massive US business losses. On the other hand, it's hard to see what any non-nuclear opponent could do to the US that would be more damaging than its own opiate manufacturers have already achieved.


> It seems possible that the US will get itself into a conflict, possibly (hopefully!) non-total, that disrupts its globalised supply chains and results in massive US business losses.

Before WW1 happened, people thought there was no way a major war in Europe, entangling the major powers, would happen because all of their economies were tied together and it would result in a major economic collapse. However, the war still happened.


One "good" thing about it is that if they are stupid enough to do it anyway the damage means they are less likely to be capable of repeating their mistakes with the same magnitude.

Granted the circumstances can still set the stage afterwards for future conflicts even if the fools responsible lost power as unfortunately demonstrated by WW2.


That McDonald’s thing sounds clever but it boils down to “the US’s Cold War system of alliances lined up closely with its economic ties.”

Were there no McDonald's in Iraq? Although that exception doesn't disprove the underlying point; Iraq is probably the most developed country America has been at war with since the Second World War.

Well, it would be pretty easy to come up with a lot of counterexamples, principally because neighbors are both more likely to trade with one another, and more likely to go to war with one another. Every colonial war for independence was a war between two entities that were doing a lot of trade (the purpose of the colonial system). The entire colonial empire era was based on making war to enable trade, including the various European wars in China to force them to trade more. Japan was forced to trade with the U.S. by "gunboat diplomacy". The U.S. Carter Doctrine says that we will go to war if anyone tries to prevent us trading for oil.

Of course, there are counter-arguments to each of these examples, but that's kind of the point: trade and war don't happen on a random basis, so which one is causing or preventing the other is hard to say. Nations that expect to go to war soon, start looking for other trading partners. It doesn't mean that continuing the trade would have kept the war from happening, it just means wars are usually preceded by a period of rising animosity short of war.

Plus, the Ukraine's decision to pursue EU membership did not lead to a reduction of the threat of war in the Ukraine. Quite the opposite, actually.


Retaliation / MAD is a non trivial idea that leaders of nations should understand.

I only worry of a cornered politician or spoiled Heir that would ignore this.


That line of thinking didn't stop World War I from happening[1].

[1] - https://www.washingtonpost.com/news/monkey-cage/wp/2014/04/0...

"Political scientists often argue that international economic relationships can decrease the likelihood that states will engage in war. Nations that might otherwise be inclined to fight can be deterred, informed or transformed by economic interdependence.

World War I has been called the “Achilles Heel” of such theories. In the decades before the war, Europe experienced unprecedented growth in international economic interdependence, both through trade and capital flows. Yet war not only broke out, but the scale of its destruction was unprecedented. Economic liberalism went into retreat. How could anyone sustain the notion that trade could prevent war when the most destructive war in history had followed an era of expansive international trade? For decades, scholars and policymakers alike pointed to World War I as providing strong evidence against the liberal case."


As I recall it, Barbara Tuchman argues/describes in the Excellent "The Guns of August" that once precautionary mobilization had been ordered in one country, due to the nature of the railway system driven technology behind the logistics of the day, war was pretty much inevitable.

Nobody realized "the system" had this side effect before it was started, and once the button was pressed, it was pressed...

As your quote says. "economic relationships can decrease the likelihood that states will engage in war".

https://smile.amazon.com/Guns-August-Pulitzer-Prize-Winning-...


But was this really in the nature of the technology, or was it a choice on the part of the German planners? Choosing to only have an invasion plan, and not bother scheduling out the 80%-option which leaves not quite as many troops but all on your side of the border, and so on down? That sounds like the generals wishing for war, to me, and limiting their masters' options.

Not quite parallel, but IIRC it emerged afterwards that the Soviets had no non-nuclear battle plans for Germany. While the US planners kept two sets ready.


Well, the nature of the technology certainly added to it. Once the trains were running, troops were flooding to the borders, and if stopped, they couldn't (somehow...) be restarted easily, because all the plans were interlocked so precisely. Or something.

Or that's the picture Tuchman paints.

Not just on the German side, BTW, but everywhere.

Now, this is just a popular history book I recall from memory several decades ago, and the questions you ask sound very relevant!


The article you linked says the opposite of what you quote.

Right after the quote, the article's authors say that these claims [about Europe being economically interdependent before WW1] are exaggerated and this do not disprove the theory that interdependence prevents war.


By that same argument, murder laws are ineffective because they don't stop serial killers. And traffic laws are ineffective because they didn't stop a big pile up that one time. Earthquake-safe construction is ineffective because it didn't protect against the Big One.

You can't point at severity as an argument against a mechanism to control likeliness.


From what I remember there was a period of rapid (although not total) deintegration leading up to WW1 and WW2.

When goods and services don't cross borders, soldiers will.

It's interesting to see videos of the borders of the world over time. Contrary to intuition, the rate of shifts (often due to warfare) were accelerating over the centuries. Then, suddenly, since 1945 everything comes to a standstill among developed nations excepting the relatively peaceful internal collapse of some communist nations. 1945 of course being when we first detonated a nuclear weapon over another country, twice.

The thing that's stopped countries from war has not been trade, amicability, social evolution or anything besides nukes. The reason no developed nations ever go to war is because there is no win condition. If one side or the other is facing significant losses, they will retaliate with nuclear weapons. If that doesn't end the war immediately - it'll shortly result in a flurry with both countries effectively ending each other instead. Mutually assured destruction, as a term, always felt like it somehow missed some of this nuance!

The thing that ought actually concern people is the development of anti-missile technology which is rapidly advancing. Highly effective anti-missile technology would signal the return to war. This is why countries such as Russia have been developing things like 'Tsunami Weapons' [1] - a nuclear weapon designed to be detonated under the ocean that could generate radioactive tsunamis thousands of feet high. If that sounds difficult to believe consider that we have nukes in excess of 100 megatons now a days. Hiroshima was 16 kilotons - in other words about 6,250 times more powerful.

The point of this being is that I think this will help create a multipolar world and industry - imagine e.g. a new Chinese OS competing against Android/Apple. But I do not think that the current status quo of the world has been playing a meaningful role in deterring war anymore so than a change to a more multipolar world will advance it. I think it's simple. When we can't overcome nukes, we won't have "outright" war - creating an allowance for proxy wars like Vietnam where no nuclear participant's sovereignty is at stake. When we manage to overcome nukes, we will have outright war.

[1] - https://www.rferl.org/a/russia-launches-new-submarine-believ...


Sort of. The social context has changed completely, though. In WWI war was still considered a glorious adventure for which the elites were particularly suited. New technologies like the machine gun, tanks, and poison gas would make sure any conflict would be quickly resolved. Given the glorification of conflict and woefully inadequate knowledge of how much risks had increased going to war seemed to many Europeans at the time like the most obvious and glorious choice.

Then came the horror of it all and Europe's upper classes had their youngest and brightest massacred and values changed. Warriors were now ordinary suckers and the elites kept themselves above the churn. War became a business of calculated leverage. It on that massive change in context that we changed together to deciding that war is not always such a good idea.

Then on top of that conflicts like those Vietnam and Afghanistan merely emphasize the baseness and poor value return of armed conflict. What was a sure thing before WWI was a daring bet in the nuclear age an nukes were only one factor involved in that calculation.


War didn't end following nukes or WW2 - only overt war between nations with nukes ended. The Cold War is the best example of this. The one and only reason the Cold War was not World War 3 was because of nukes. If places like Pakistan and North Korea were not nuclear powers - they would have long since been Libya'd.

> If that sounds difficult to believe consider that we have nukes in excess of 100 megatons

No we don’t. The biggest ever was Tsar Bomba, which was 50 Megatons with the option to make it 100. The aircraft that dropped it barely got out of the way in time, and it was only on a plane at all because it was too big and heavy for a missile.


Your argument is what? The largest nuke ever tested was 50MT. It is entirely possible (abd probable) that development continued. Was there a physical limit about the tsar Bomba? I think the dial-a-yield testing at only half strength is a pretty solid indicator no.

Do you think they stopped because dropping the Tsar Bomba at full yield was a suicide mission? Yes, yes it is possible that it would have been. That would not have stopped development. Suicide missions happen. We have autonomous planes now.

There's only mild speculation in the >100MT number. They almost certainly exist, and your counterexample is right on the line.


I think you have a point about nuclear bombs keeping the peace.

Exactly what one-offs the russians once made, I don't know how well this is known. But we have a pretty accurate picture of the stockpile of bombs existing now. And the the age of maximally-huge bombs was a long time ago, to compensate for terrible missile accuracy.


I'm unsure of how accurate our classified knowledge is, but our publicly stated knowledge is not especially accurate. Many countries continue to openly develop nuclear weapons such as India and Pakistan. Other countries' ongoing development is an open-secret, such as Israel. And everything that we publicly know even about those developments tends to be based on a pretty wide range of estimates. This is further compounded since countries tend to stop testing once a certain level of competence is reached.

On top of this there is undoubtedly genuinely secretive developments. I think it's improbable that all nations who signed the nonproliferation treaty have not chosen to develop nukes. This comes in two forms. The first is larger nuclear powers such as USA/Russia/China not expanding or modernizing their arsenals. The second comes in smaller nations not creating their own. For instance South Africa, during its apartheid years, managed to secretly develop nuclear weapons which it subsequently dismantled before handing power over to the new non-apartheid government. Another possible, if not likely scenario, is entrenched nuclear powers aiding in the development of nuclear capability of smaller allied nations as a means of ensuring those nations sovereignty and as a solidification of their relationships. As an example of this, the US has for decades been alleging that China has been helping to enhance and expand Pakistan's nuclear program. In times past it was alleged that Israel had played a role in the development of South Africa's nuclear program.

Tsunami weapons are also a game changer. They don't suffer from the same issues of delivery that land/air based weapons do and may be more able to effectively sidestep missile defense systems, and of course also dependent on absolutely enormous payloads.


We have a decent idea what small players like Pakistan/NK have got, I think, in terms of number & sophistication.

While I'm very doubtful that the US or Russian stockpile contains a > 100MT warhead, I agree that if there was any serious reason to produce one, this need not take very long, nor attract much attention.

Large bombs are in the ballpark of earthquakes, for energy released. But how much of this would be contained in the water, not just blown into space, or something? Presumably someone has done the calculations...


> For instance South Africa, during its apartheid years, managed to secretly develop nuclear weapons

As I understand, “develop” is a bit generous; it was largely direct transfer from Israel.

> Another possible, if not likely scenario, is entrenched nuclear powers aiding in the development of nuclear capability of smaller allied nations

How is that not likely? Most current (and the few past) nuclear powers that aren't the US, Russia, or perhaps Israel benefited from this at some point in their programs, covertly or overtly.


The Tsar Bomba is the most powerful weapon ever detonated, and that was in 1961. Interestingly even that weapon back then had a 100 megaton design. They decided to scale back the tested payload for a variety of reasons including enormous fallout and that testing such a massive weapon would be a suicide mission for the pilots dropping it.

Should also probably add that it was so physically massive the bomber's interior and Bombbay doors had to be stripped, it required a parachute that itself weighed around a ton, and yet proved to be tactically underwhelming for its cost because the fireball ended up blasting itself into the upper atmosphere before it had time to fully expand

In many cases, and even more so in the most critical components, there is no second tier of suppliers. There is no market for second tier components. Using second tier components would subtract so much value that OEMs in highly competitive businesses, like Apple or Samsung, can't even consider such substitutions.

So we can't "go it alone." That may be a feature, not a bug.


Hackernews darling Backblaze talks about how they were affected when the hard drive shortage happened due to the flooding in Thailand back in '11, which is how we found out 50% of all of the hard drives in the world are produced there.

Sourcing high value MLCC capacitors is a real problem. I often design something into my board, only to find out it has gone out of stock. It’s like playing whack-a-mole.

Good high level history of MLCC (including graph from 2017 showing shortages): https://www.ttiinc.com/content/ttiinc/en/resources/marketeye...

Looks like the high capacitance are the longest lead, which is what I have been finding. 10 uF and up in 0603 are hard to get.

We made an item in our design checklist last year to avoid high value MLCC caps whenever possible. We also made sure that no designer can add new high value MLCCs to the component library :)

> Without programs like the Linux operating system or Kubernetes, a tool to manage computing loads, Alibaba could not have become the world’s fastest-growing cloud-computing giant

Kubernetes i could probably understand, but Linux under export control?





Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: