Tesla investigated for flash memory failures
-
-
D'OH!
Stock on its way back to 420?
-
@just-jeepin Whew. Good news the gauges are indicators are there as redundant backups....
-
Kwality cars!
-
Yet people continue to buy them. I will NEVER understand the brand loyalty and the droves of idiots who look all their problems and say, "Yes, I want to pay tens of thousands of dollars to deal with a shitty car."
-
@his_stigness said in Tesla investigated for flash memory failures:
Yet people continue to buy them. I will NEVER understand the brand loyalty and the droves of idiots who look all their problems and say, "Yes, I want to pay tens of thousands of dollars to deal with a shitty car."
Cult-like following isn't exactly rare around this country lol
-
@his_stigness 50% of people are of below average intelligence
-
@just-jeepin All high density flash should be serviceable. That shit's fragile.
-
Tesla has a history of this type of problem. They want to do things differently, so they avoid hiring to many experienced automotive engineers. In many cases, this works out well. Tesla has found some innovative ways of making an electric car. There are good reasons not to do things the way they have always been done in the past.
On the other hand, this means that you sometimes make stupid mistakes. There can be an attitude of "We are doing things in a way that haven't been done before!". Sometimes it wasn't done before because nobody thought of it. Sometimes it was because other people did think of it, but realized it was a bad idea.
For example, there were the sensors on the door handles for the Model S. Tesla though they could save a lot of money using solid wire to the sensors instead of stranded wire. However, that wire had to bend slightly every time the door handle presented/extended. The first fix was to move to stranded wire, but there wasn't anything to seal out moisture at the ends. Finally, they figured out how to use flexible printed circuits and waterproof coatings. The older auto companies learned not to use solid wires for flexible connection nearly a century ago. Tesla had to learn it on its own.
It's likely the same situation with this memory. An experience engineer would know not to use it for something that gets written so often, but Tesla has to learn it on their own.
I know a lot of Tesla owners and most of them like the newest technology toys. They didn't get their Tesla because they wanted a good car. They got it because they wanted a high-tech car. They think of themselves as early adopters and they expect bugs. They are willing to accept these problems as the cost of having the latest technology.
-
Not sure what bothers me more- that this tech struggles to last even ten years, or that such critical vehicle functions are dependent on it.
-
I did some quick back of the envelope math and 3000 cycles would be equivalent to 2 years of vehicle use in my household.
-
@roadkilled said in Tesla investigated for flash memory failures:
Tesla has a history of this type of problem. They want to do things differently, so they avoid hiring to many experienced automotive engineers. In many cases, this works out well. Tesla has found some innovative ways of making an electric car. There are good reasons not to do things the way they have always been done in the past.
On the other hand, this means that you sometimes make stupid mistakes. There can be an attitude of "We are doing things in a way that haven't been done before!". Sometimes it wasn't done before because nobody thought of it. Sometimes it was because other people did think of it, but realized it was a bad idea.
For example, there were the sensors on the door handles for the Model S. Tesla though they could save a lot of money using solid wire to the sensors instead of stranded wire. However, that wire had to bend slightly every time the door handle presented/extended. The first fix was to move to stranded wire, but there wasn't anything to seal out moisture at the ends. Finally, they figured out how to use flexible printed circuits and waterproof coatings. The older auto companies learned not to use solid wires for flexible connection nearly a century ago. Tesla had to learn it on its own.
It's likely the same situation with this memory. An experience engineer would know not to use it for something that gets written so often, but Tesla has to learn it on their own.
I know a lot of Tesla owners and most of them like the newest technology toys. They didn't get their Tesla because they wanted a good car. They got it because they wanted a high-tech car. They think of themselves as early adopters and they expect bugs. They are willing to accept these problems as the cost of having the latest technology.
This fault has nothing to do with cars and everything to do with the service life of flash storage. People working at a company like Tesla should know these things. This was either a cost cutting measure (and a "let's deal with it when it happens" type situation), a crunch time mistake or just gross incompetence. Maybe even a combo of those three.
So this is either a massive fuckup by Tesla or intentional dishonesty and neither are acceptable. I mean if we're going to rake car companies over the coals for their screw ups, Tesla should get the same treatment.
-
@poor_sh said in Tesla investigated for flash memory failures:
@his_stigness said in Tesla investigated for flash memory failures:
Yet people continue to buy them. I will NEVER understand the brand loyalty and the droves of idiots who look all their problems and say, "Yes, I want to pay tens of thousands of dollars to deal with a shitty car."
Cult-like following isn't exactly rare around this country lol
No idea what gave you that impression.
-
@bloody-the-resident-shitposting-saffer said in Tesla investigated for flash memory failures:
@roadkilled said in Tesla investigated for flash memory failures:
Tesla has a history of this type of problem. They want to do things differently, so they avoid hiring to many experienced automotive engineers. In many cases, this works out well. Tesla has found some innovative ways of making an electric car. There are good reasons not to do things the way they have always been done in the past.
On the other hand, this means that you sometimes make stupid mistakes. There can be an attitude of "We are doing things in a way that haven't been done before!". Sometimes it wasn't done before because nobody thought of it. Sometimes it was because other people did think of it, but realized it was a bad idea.
For example, there were the sensors on the door handles for the Model S. Tesla though they could save a lot of money using solid wire to the sensors instead of stranded wire. However, that wire had to bend slightly every time the door handle presented/extended. The first fix was to move to stranded wire, but there wasn't anything to seal out moisture at the ends. Finally, they figured out how to use flexible printed circuits and waterproof coatings. The older auto companies learned not to use solid wires for flexible connection nearly a century ago. Tesla had to learn it on its own.
It's likely the same situation with this memory. An experience engineer would know not to use it for something that gets written so often, but Tesla has to learn it on their own.
I know a lot of Tesla owners and most of them like the newest technology toys. They didn't get their Tesla because they wanted a good car. They got it because they wanted a high-tech car. They think of themselves as early adopters and they expect bugs. They are willing to accept these problems as the cost of having the latest technology.
This fault has nothing to do with cars and everything to do with the service life of flash storage. People working at a company like Tesla should know these things. This was either a cost cutting measure (and a "let's deal with it when it happens" type situation), a crunch time mistake or just gross incompetence. Maybe even a combo of those three.
So this is either a massive fuckup by Tesla or intentional dishonesty and neither are acceptable. I mean if we're going to rake car companies over the coals for their screw ups, Tesla should get the same treatment.
Flash storage can be ridiculously reliable...look at current Solid State hard drives for computers....so they likely also went with really shitty quality components.
-
It's kind of sad you can't tell with Tesla whether it was an intentional cost cutting measure, or a fuck up.
Either way, they look terrible. But it's worse you can't easily guess which it was.
-
@rallydarkstrike As someone who works with 30 datacenters, yeah it's pretty damn reliable if you don't buy a shitty $20 SSD.
-
@bloody-the-resident-shitposting-saffer said in Tesla investigated for flash memory failures:
@rallydarkstrike As someone who works with 30 datacenters, yeah it's pretty damn reliable if you don't buy a shitty $20 SSD.
Yup, agreed! I only have Crucial MX500-series and Samsung EVO 850/860 series in my personal builds!
-
@mybirdistheword 50% are below mean intelligence but I have a feeling FAR more than 50% are below average intelligence ...
-
I've never heard of a flash memory that can't handle a minimum of 100,000 re-writes. Sounds like they cut the wrong corner
-
@frinesi2 average an mean are basically the same thing lol
-
What's to investigate? This is a well-known Tesla fuck-up and general "rookie design mistake".
This is also a side result of the "Move Fast and Break Stuff" design mentality-- which means, like a GM product on steroids, you never want to own the first few model year runs of a Tesla product.
Between the non-working doors and MCU issues on an early Model S, the design nightmare of the X FulkinWing doors or the detachable-bumper-in-rain on the Model 3? Forget the first 36-48 months of production--- those are all "engineering mules".
-
No. Mean and Median are two different forms of "average"
-
There's a well known limit on R/W cycles. The embedded OS in SSDs manages that limit.
This was all rookie, all-unforced errors at The Big Lazy T
-
@mybirdistheword I meant median ...
Well, I guess I know where I stand now ...
-
@his_stigness said in Tesla investigated for flash memory failures:
Yet people continue to buy them. I will NEVER understand the brand loyalty and the droves of idiots who look all their problems and say, "Yes, I want to pay tens of thousands of dollars to deal with a shitty car."
It's not unusual. BMW and Land Rover and Jeep all have very dedicated buyers.
But aren't we all here on Oppo because we value certain aspects of a car more than others? How many people here would choose a reliable Camry over anything else on the market? Hell, how many of us would choose a Model 3 instead?
Anyways, considering that any Tesla has no head gaskets to fail, no DCT transmissions that shift horribly, no emissions they can cheat, not even an ignition switch that can fail.... Well, they can afford a few failures elsewhere before owner satisfaction can dip to the level of some other manufactures. (Though they certainly seem to be exploring the boundaries of that "head start")
Overall though, I'll stick to my theory: Reliability is overrated.