> 5G doesn't require imploding plutonium.

Reading HN, I was under the impression that we didn't need 5G anyway (except for people streaming videos in stadium, which I haven't really understood why yet).

5G's very high bandwidth and other features help to solve the following issue: when you get a huge population close together all communicating, (as in stadiums, conventions or during emergencies in very dense population centers) the airwaves become too congested to get more than a few KB/s with older tech.

Or even just in dense cities. I see 4G speeds as high as 90 Mbps at (eg) 6AM, but at 9PM on a weeknight it can slow to 2 Mbps or less due to congestion. Too slow to stream video with decent quality or to download anything, and it’s very frustrating!

I can stream Netflix on my 25mbit/s. For personal use I absolutely don't care about 5G. For professional use it might give me better picture quality when I remote into work, but I think my work's VPN server is more of a bottleneck currently.

> ”I can stream Netflix on my 25mbit/s. For personal use I absolutely don't care about 5G.”

I can stream at 25 Mbps just fine on 4G+/LTE too. At off-peak times. I’ve even seen it hit 70, 80, 90 Mbps occasionally.

The problem happens at peak times, usually starting around 6pm, when the network slows to a crawl and I can’t stream things reliably despite having 5/5 signal strength, because my carrier (Three UK) has sold too many cheap unlimited data plans and their 4G network is highly congested.

5G is needed because it greatly improves the data carrying capacity of given blocks of spectrum, meaning more users and higher speeds in areas that currently suffer capacity constraints.

And honestly, 5G’s 200 Mbit+ speeds would just be nice to have to speed up things like downloading multi-GB OS updates and the like.

That might be less to do with your carrier's cellular edge capacity (are you in a densely populated area with lots of people on the same cell station as you?) and more to do with your carrier's intenet edge capacity (3 has lots of customers using a lot of internet)

If it's the former, 5G helps you. If it's the latter, 5G makes zero difference to you.

Yes, I can’t say for sure but I assume it’s the former. It’s certainly a dense area (Central London) and there are certainly many people around here with 3 unlimited data plans.

If I monitor my router’s status page I can see it start to bounce between several different Cell IDs when the congestion hits, and the SINR drops to low single digits (from 16-20dB when it’s fastest).

In any case, as I understand it, backhaul upgrades will be a part of 3’s 5G rollout anyway.

I tried 5G in Shenzhen on a demo phone, still only 10 mb/s to Hongkong, but you can have 300mb/s to the nearest NX.

Unless your server sits right beside the cell tower, you will never see that promised 1gb/s

Is that much of an improvement on 4G? I often see 250 mbps on LTE during off-peak hours

Your internet speed and latency may be fast enough for the current media/mediums on the internet, but I think it would be naive to assume (without strong evidence/reasoning) that "there's nothing left to be invented" which might consume more bandwidth, or require lower latencies in the near future.

Reading some of the comments I see that I don't have the same use case (data cap, so never streaming video ; and rarely need to vpn or do remote work with only a mobile connection available).

I am excluding future usage though (biggest I could see is "That smartphone is your only mean of connection to the net, either via the device or when acting as a modem.").

This. Especially because 4G is far, from being deployed everywhere.

Well if it made VOIP a great experience instead of an so-patchy-you-have-to-survey-users-after-every-call experience then it might be a big deal.

This requires a supply chain that the US no longer has, AFAIK. I mean, do know of any electronics still made in the USA? Shenzen is the world's 3D printer, and its non-trivial to clone it.

Yes, lack of manufacturing capacity in western nations is the root of the problem, not the particular technological fetish de jour. This is the chickens coming home to roost of chasing shareholder value maximisation rather than keeping some spare capacity in the system for a rainy day.

China is not the enemy, we should work with them to build a healthy middle class existence for all the world's citizens. Having one a country mono-culture doing the lions share of the advanced manufacturing is just asking for another dark age when that society suffers a catastrophe.

I’d like to believe this vision of superpowers working together for the betterment of humanity. But I feel like the moment anyone gets a slight advantage they leverage that to squeeze everyone else - economically, technologically, fiscally, geopolitically, you name it.

As is having one country doing the lion’s share of advanced software. No?

> Yes, lack of manufacturing capacity in western nations is the root of the problem

The reality here is this quickly leads to a discussion about regulations some people would be furious at the even the thought of re-examining.

> This requires a supply chain that the US no longer has, AFAIK....and its non-trivial to clone it.

The US (and other countries) lost much of its electronics supply chain (and manufacturing in general) due to political priorities, and it can regain it by shifting those priorities and enacting new policy. This idea of "it's gone and never coming back" is false and a result of an over-attachment to certain ideas and certain policies.

>it can regain it by shifting those priorities and enacting new policy

that would require a fundamental rethinking of economics and trade away from free-marketeering towards an understanding more similar to the Chinese who put these questions into the context of national (security) interest, and that is very unlikely to happen, in particular in the US in which private interest groups wield significant power over a largely weak and incoherent government. In Europe, it may be possible if France would have more of a say but countries like Germany or the Netherlands would be significantly hurt by the inevitable retaliation so it's probably a nonstarter here too.

It feels to me like the US as a whole is getting more protectionist over the last decade, and that the free trade everyone is welcome, export to our market for free is seeing some massive attacks.

This is a feeling shared by both the left and right in the populist portions of both parties.

People are looking for ways to get more solid middle class jobs back into the USA, and protectionist trade policy is an obvious place to start for many.

Prior to Bretton Woods there had never been a system of free trade between (nearly) all countries enforced by a singular superpower.

There's nothing fundamental saying that America won't swerve away from providing the service of protecting and enforcing global trade. If that ever happens, it will have many serious consequences, but the US could still decide it is in its best interest.

> ...in the US in which private interest groups wield significant power...

AFAIK, those interest groups don't actually wield very much power, but they've been very effective at promoting the fatalistic attitude that their favored polices are inevitable and permanent.

For proof, look at the current president who started a major tariff war that was very much against a major leg of free-market policy. What paused it wasn't those interest groups, it was political uncertainty caused by retaliation that targeted farm states by the other party in that war.

Unless the manufacturing industry was heavily subsidized I doubt it could compete with the Chinese.

it is not just political priorities. either the onshored industries would have to be as competitive as those in asia and europe, or they would have be subsidized. these are not trivial matters.

Or they could be neither but protected by tariffs or other means. The decision is one about political priorities, in this case: free trade vs. other considerations.

Arduino was originally Italian, made in EU. There are modern power tools, car electronics, and coffee machines made in Hungary - I'm only listing the ones I know for sure. The components however - IC, resistor, capacitor, etc - are nearly all from China.

Semiconductor fabs, especially for high-tech stuff, still are somewhat spread around the world at least (even though the later chain isn't, i.e. you might see a fab in Germany, and the cut wafers then being flown to Vietnam to put them in the plastic IC cases because that's cheaper).

For simpler components (resistors, capacitors) special types still have some manufacturers in the west (e.g. high-precision, especially robust variations, ...). Cheap mass-production not so much.

I remember trying to find out if you can get EU-made LEDs, and the result was that maybe some high-power ones by OSRAM would fit the bill, but it wasn't clear what was made where.

You can get LEDs from Taiwan, Malaysia, Japan. You just don't get the same level of quality that you get from top-end Chinese manufacturers that way. Similarly passives are often made in Mexico (to bypass US tariffs), Vietnam, Malaysia again, Thailand. But the most consistent quality product at large quantities still comes from China.

Japanese LEDs, eg Nichia seem to be pretty high quality: https://people.xiph.org/~xiphmont/thinkpad/led-gamut.shtml

Yep, Nichia LEDs are excellent. But I prefer Chinese manufacturers for LEDs - they're much easier to deal with and quality is good and consistent and availability is great. I recently had some LED emitters custom made in China (low quantity) and the light properties are exactly to spec. I wanted a combination of cri and color temp that wasn't available off the shelf in that form factor. At the quantities I wanted, nobody outside China would even talk to me, but I had several options to choose from in China and the first one I talked to delivered, exactly to spec, exactly on time.

Japan still has components if I'm not mistaken, LEDs especially.

> I mean, do know of any electronics still made in the USA?

Most Intel chips are still being made in the USA. https://en.wikipedia.org/wiki/List_of_Intel_manufacturing_si...

This shift is largely due to the cost of human capital. If automation advances exponentially like many believe it will, should we expect much of this manufacturing to move back to the US? At that point, shipping would probably be the largest variable cost in getting goods to market.

No, shipping is such a small component of most product costs that it can never have a huge effect. (Except perhaps for heavy, bulky low value goods.)

My understanding is that the Huawei's hardware is just way cheaper.

It is. Ludicrously cheap. Not just in capital expenditure terms but also in terms of operating costs, as they’ll run it all for you, if you want. There are lots of possible reasons ranging from business pragmatics, to work practices through to the tin foil hatted. As far fetched as any of these might seem, none are as far fetched as that they’re giving you a great deal out of pure kindess.

Take note that the ludicrously cheap prices only apply if its end-to-end Hauwei: i.e. If you plug pne piece of Hauwei kit into another.

If you add non-Hauwei kit, you not only have to buy the more expensive gear but you also have to buy a more expensive licence for the protocol on the Huawei kit.

Source: adding SMSCs, GMLCs, SMLCs, etc to HW networks.

Yes I work in this space too, and as a 3P vendor their monolothic architecture concerns me greatly. Note that this approach goes directly against the spirit of the GSM “culture” (for want of a better term) of modularity to date.

It's more like they bought it from us. ARM used to be owned by Westerners. https://www.wsj.com/articles/softbanks-arm-holdings-to-sell-...

Nokia is currently extremely cheap. P/S 0.85 and it includes Bell Labs, and Lucent and Motorola wireless networks assets in the US.

It's really not about owning.