Chip aging becomes design problem (2018)

My take: for as long as recycling performance is terrible, this should be a no no. Social movements should start demanding better products.

Many components in your consumer devices age; my greatest concern is electrolytic capacitors (specifically the tantalum ones). I think many game consoles and computers with switched-mode power supplies are unlikely to last decades. If you want a simple way to get your devices to last longer, I suggest that you pick ones with external (brick) power supplies, as the SMPS is likely to be one major cause of failures.

>"I’d much rather we over-design stuff to last decades, at least at the chip level where overedesigning is super cheap."

I am not sure that people are willing to pay significantly more for these longer-lasting devices, and all that extra effort will be wasted if the devices are scrapped prematurely. It may be more (environmentally) efficient to simply replace the devices upon failure rather than over-designing them.

Unfortunately, "we" aren't really considered in the decisions. The math from a company's perspective is mostly, "Will this last the expected warranty period?" And, to a lesser extent, "Will the ill will caused by these catastrophically failing right outside warranty be a problem?" See RRoD and such.

But the problem is that "solid state" does wear out - this article is a list of causes of it. It's reasonably true that the older solid state technologies lasted extremely long, but as you start to push them, they don't last as long, and even things like power transistors eventually start to fail - there are a few companies that rebuild old Tesla Roadster power conversion equipment, because the transistors wear out and fail (at least one of those companies seems to regularly burn down their shop as well).

> It would suck so bad if we nickel and dimed what should be fundamentally physically robust devices to last for much less time than complicated mechanical devices from the past.

For you, sure. For the people in business selling replacement, a widget that lasts 30 years is an annoying pain in the ass to them.

GMC used to build transit buses (beautiful aluminum chassis, slanted windows, the works), and eventually stopped. Talking to the bus mechanics when I used to drive, their theory was that GMC couldn't sell new ones, because damned near every single one they'd made was still on the road. They didn't corrode, and a properly maintained chassis would last basically forever. In the early 2000s, we had 40 year old buses that were on their 4th engine, 3rd transmission, up near a million miles. They just kept trucking along.

Now, it seems consumer stuff is lucky to last 5 years before needing major repairs.

I remember the same being said about LEDs (and before that, CFLs) for lighting, that they would have a much longer life, but in practice, they didn't precisely because making something that lasts longer means less profit in the long term.

Also, the aging calculations are done under worst possible conditions, ie the computation is for 2years active life for a consumer device, at highest operating voltage, highest operating frequency and highest operating temperature. The time while it is powered down or sleeping isn't aging the device and so doesn't impact lifetime. The actual working life in your hands under regular usage will be (much) longer. There is a reason your average phone works fine for 5+ years from hardware perspective, even if the software side sometimes has shorter planned obsolescence (lack of updates, or updates that demand more from the device than that generation of chip was capable of delivering, etc)

For telecommunications, or automotive usage, and in particular for anything that might be failure critical, the testing and metrics are much more onerous, and the minimum guaranteed lifetimes higher. This partly explains why such devices, even when manufactured at advanced nodes, don't exhibit the same performance as consumer devices: some of the performance is being held in reserve to account for the natural degradation due to aging.

Source: I'm one of the people quoted in the original article.

Sorry to be cynical, but says who? The marketing department? Product lifecycle is a function of engineering design, and not necessarily material choice. A Detroit Diesel engine will run basically forever, yet I would be surprised if many of the newest electronic "no moving parts" widgets will last more than a few years. If fewer moving parts is inherently more reliable, then ask an EV owner why charging points are so frequently broken. After all, they only have one moving part!

The last few days before the Anti Singularity will be terrible. We will be desperately trying to complete the next generation of engineering designs, while our current systems crumble and age before our very eyes.

Engineers are working on it.

From a follow on article by the same author.

"An emerging alternative is to build aging sensors into the chip. “There are sensors, which usually contain a timing loop, and they will warn you when it takes longer for the electrons to go around a loop,” says Arteris IP’s Shuler. “There is also a concept called canary cells, where these are meant to die prematurely compared to a standard transistor. This can tell you that aging is impacting the chip. What you are trying to do is to get predictive information that the chip is going to die. In some cases, they are taking the information from those sensors, getting that off chip, throwing it into big database and running AI algorithms to try to do predictive work.”

https://semiengineering.com/aging-problems-at-5nm-and-below/

For other hardware, there is almost the of redundancy. In those parts of the system, you depend on multiple components all working correctly with no chance of detecting errors at the circuit level. This means that if you have 10 components with expected life of 6-20 years (with a mean of 10) you can expect an actual life of about 6, not the mean life at all. The weakest link and all that.