This mimics numbers from Denmark, where Model 3 had a failure rate of 20-25% after four years a couple of years ago, and last year the first Model Y's had to go through their first 4 year inspection and 45% failed. 34% of Model 3's failed last year.
For comparison: Last year VW ID4 had a failure rate of 2%, and the average for _all_ electric cars (no matter age, including Teslas) was 7% failure.
Causes: Breaks, wheels, steering, and a few more critical things along those lines.
Objectively speaking, Tesla cannot manufacture cars that live up to European standards.
The report from the original article is not a general problem rate but more specific: TÜV does mandatory technical inspections every two years. In those inspections, only safety- and environment-critical problems are checked for, so e.g. brakes, rust on structural parts, high emissions, non-working lights. But there is a whole bunch of stuff that they don't check for, e.g. heating/cooling, GPS not working, doesn't charge/start sometimes, ...
So it's quite possible that both are true: Maybe ID4 has lots of non-safety and non-environment problems, so it is in the shop very often, but still rarely fails an official inspection.
A sample of one but ours did fail the inspection (suspension). It also experienced a complete shutdown of instrument panel on the motorway: not something you reproduce easily in a regular inspection but a pretty damn serious condition. Fail to unfold the mirrors/engage parking assistant or rearview camera happened dozens of times.
None of other owners I spoke to were particularly happy with theirs either.
I also experienced a shutdown and restart on an one-week-old Opel (but happened a decade ago). In one second everything was back online and working, but boy the adrenaline kick you get form that...
Assuming you are not an outlier, could it be VW has a low TÜV failure rate because they are in the shop often?
I have no idea what German auto shops do, but whenever I take my car in to a shop in the US for service (routine or otherwise) they generally include various inspections and adjustments to various things, including things that Google is telling me are part of the TÜV inspection.
Yes, but if a car is using regenerative braking 99% of the time, the car should track this and use brakes occasionally to "polish and maintain" them. It's not hard, and if the pads are running out, it should warn the user. Tesla does neither AFAIK.
You should check your tires, yes. At least while changing from winter to summer and vice versa, however if the cars torque profile is too aggressive and it's eating tires, you should note it at the user's manual that thread wear should be checked more frequently with respect to other cars.
> how is that to do with Tesla manufacturing standard?
My friend's Toyota Auris needs new discs every 100,000KM, new pads every 60,000KM. I change discs around 60,000KM (heavier car, mostly rush-hour traffic, hilly city, automatic transmission), and never failed an inspection w.r.t. braking power.
> how is that to do with Tesla manufacturing standard?
Unless further data/evidence is provided, it is reasonable to assume all car owners treat their cars equally shitty, and as such can be ignored in this equation since it applies equally to all manufacturers.
Exactly. I don't understand the focus on VW here. That wasn't the point of my original post at all.
Tesla didn't even recognize the inspection failures in Denmark as real at first, so it's probably fair to assume that they're only now trying to sort out the problems on new cars, and that we'll see many more failing Tesla inspections the coming years, even on cars sold up to this day.
Worn brake disks are a manufacturing problem. Nominally my VW needs new brake disks every 100Mm. Practically it needs new ones every 40Mm, because VW makes them from shitty steel that rusts and wears like hell, especially when there is salt on the roads in winter.
Some manufacturers use better steel and therefore have a longer disk lifetime.
I do drive to work almost every day, and I don't drive an electric car. So there is sufficient use.
And quite a few decades ago, people noticed that when you mix chromium, nickel, vanadium or things like that into your steel, it doesn't rust. Car manufacturers are just very slow in noticing.
That's interesting with the pre-inspection. I haven't heard about a systematic pre-inspection here. I also don't think it really matters, the most important metric I'm quoting is 7% failure rate across _all_ electric cars, and no way that's caused by every non-Tesla owner going to a pre-inspection.
(inspection costs around 80 euros in Denmark, so there's no financial reason to go to a pre-inspection anyways, just do the inspection and have it redone if the car fails).
Tesla wouldn't even recognize the problem at first, and refused repair of customer cars. Of course there's issues with every brand of car. It's just that the numbers show that Teslas are much, much worse with regards to safety critical components.
You bring your car in for yearly maintenance. They do a 50 point inspection, fix what is worn and replace fluids where necessary. After that they bring it to a shop next door where they do the (government required) yearly technical inspection. Nearly all brands do this as it's easy money and because you can persuade the customer to buy a new model of car while they are already in the shop.
Tesla does not have this. People just bring it straight to the yearly technical inspection.
Teslas doesn't fail because of fluids or worn brakes. They failed due to causes that Tesla didn't even recognize, as I already told you, because they thought wobbly wheels were ok, and other structural issues.
When electric vehicles started to become mass produced one of the selling points was due to fewer mechanical parts, there’d be less wear and fewer parts to fail and replace on electrics; however it seems electrics have introduced other complexities that kind of wash the advantage of fewer parts…
You're ignoring the basic fact that Teslas doesn't fail because of the electric drive chain, but because of basic things like wheels, suspension and brakes. Sure, electric cars are heavier, but heavy cars (vans, trucks) have existed for ages.
The problem here is that Trump believes that the Norwegian government has any say in what a private organisation is doing, and - to be frank - just shows that Trump is a tyrant who wants everyone to use illegal force to please him.
They would be The Norwegian Nobel Committee, who at this point should realize what a disaster their prize has been, and not only last year. It was inherently poorly conceived, and shouldn't be awarded to the living, who can and do go on to wage war, agitate social instability, and act against the interests of peace.
Having a situation where a president demands the peace prize otherwise he causes a war isn’t a good look for Nobel, and shows that we’ve now moved so far from the original intention of the prize that the peace component really should be scrapped.
> isn’t a good look for Nobel, and shows that we’ve now moved so far from the original intention of the prize that the peace component really should be scrapped.
I don't follow. Are you saying the committee should have known that Trump would literally wage a military war if he isn't awarded the peace prize? Are you saying if they changed their mind now and allowed Venezuelan politician Machado to gift her prize to Trump, that Trump would no longer have a desire to own Greenland? I'm honestly trying to understand but maybe I missed an important story.
Well it should be apparent that Greenland is the sovereign territory of another NATO member, Denmark. Coming along like a transactional narcissist and claiming you “need it or else” and breaking nato over it OR you get the Nobel peace prize for not capturing it.
Right! Either scrap it, or award it only to (A) those recently deceased who have devoted their lives to making peace, or (B) defunct organizations who have completed their mission and had operated in the interests of peace.
Giving a "peace" award to living people/organizations -- who can and do go on to sully the award with most unpeaceful deeds -- is a proven failure.
> Right! Either scrap it, or award it only to (A) those recently deceased who have devoted their lives to making peace, or (B) defunct organizations who have completed their mission and had operated in the interests of peace.
The legal trust for all the Nobel Prizes state (AIUI) that they can only be awarded to living persons.
The only option would be to not award it (like happened in 1948).
As if such a pesky geographic detail ever stopped USA from invading Afghanistan when fifteen Saudi, two UAE, and one Lebanese nationals attacked them in 2001.
Can we please not kid ourselves with thoughts about how this being good from certain perspectives, when the development is _clearly_ bad for consumers?
I do agree with the negatives, but at the same time, I do see some upside. I live in a cycling city, and need to rent a car maybe once a year. why then should I bother myself with the annoyances of vehicle ownership?
Let me be clear here: I do not own a car and I live in a city that doesn't require car ownership.
There is a difference between choosing not to own something because it is personally more efficient or reasonable to do so, and being priced out of owning something. I don't own a car because I don't need it, I rent because I cannot afford a home.
Done charitably, I think the mainframe model of shared compute does meet most person's needs where they don't need to care about latency. It would allow us to take advantage of economies of scale. The problem, imo, is that no one has an incentive to do this as a service, so it would turn into rent-seeking.
Sure, a shared model does make sense in many ways. We could share within a family, neighbour cooperatives, and similar scales. With the users co-owning the means of processing.
But the current model is that we all rent from organisations that use their position of power to restrict and dictate what we can do with those machines.
But I do care about latency . . . and I want things to still work when the wifi is dodgy. I already find things like Office 360 deeply frustrating (only use it for work).
You do, but most people don't. So not enough people will complain that it will make any difference. And the people who don't complain will just keep forking out money because they're addicted.
The problem is benchmarking on the 96 vcpu server, because at that point the author seems to miss the point of Kafka. That's just a waste of money for that performance.
And if the OP hadn't done that, someone here would complain, why couldn't the OP use a larger CPU and test if Postgres performs better? Really, there is no way the OP can win here, can they?
I'm glad the OP benchmarked on the 96 vCPU server. So now I know how well Postgres performs on a large CPU. Not very well. But if the OP had done their benchmark on a low CPU, I wouldn't have learned this.
you're missing the point. Postgres performs well on large CPU. Postgres as-used by OP does not and is a waste of money. It's great that he benchmarked for a larger CPU, that's not what people are disputing, they are disputing the ridiculous conclusion.
I remember doing 900k writes/s (non-replicated) already back on kafka 0.8 with a random physical server with an old fusionio drive (says something about how long ago this was :D).
It's a fair point that if you already have a pgsql setup, and only need a few messages here and there, then pg is fine. But yeah, the 96 vcpu setup is absurd.
