From the description of events, it seems more likely that the thermal runaway already started. The CPU detected the critical heat and shut off, as it is designed to do to protect THE CPU from damage. The problem is that it wasn't the CPU creating the heat, but the battery. Turning the device back on likely had nothing to do with the problem.

I think this comes from a lack of understanding about how the problem started. Yes, excessive heat CAN cause a critical battery condition, but this seems unlikely, as the batteries should be designed to handle much more heat than the CPU can generate. Rather, I think the batteries caused the heat, not the other way around.

Except that normal CPU usage assumes proper airflow. Once the CPU shut down, he turned it back on which could have potentially put the battery back into a critical state. If the battery was already cycled out, then that would further support that story.

> The guy decided to turn it back on, which is when it blew up.

Which is exactly what anyone else would do if their laptop suddenly shut off without giving any clear warning that the situation was dangerous. A proper safety system would need to notify the user of the possibility of a battery fire and advise them to take the laptop somewhere where fire damage would be minimized.

But it's not entirely clear to me that safety features were actually triggered. Perhaps the battery failed and the machine shut down due to lack of power or a power fluctuation. The fire could have been inevitable at that point, and his actions after the shutdown and before the fire are just red herrings.

If your computer randomly turned off, wouldn't your first thought be to turn it back on?

If it did indeed recognize the heat problem and turn itself off, they should likely either tell the user or not allow it to turn back on until it cools down.

If your thermal "don't explode" protection can be overridden by pressing the power button, what's the point?

I'm not sure this is what actually happened, anyway. He says it caught fire within a few seconds of being turned back on. Either the thermal protection is faulty (it should shut off before the battery gets anywhere near that point) or the machine shut off for some other reason (excessive CPU temperature, say).

I don't know about Macbooks, but my desktop PC will refuse to POST if temps are past the critical threshold.

My lenovo laptop sometimes tests if boil-in-bag might be a good idea for laptops now even if it hasn't been anytime before.

I'll notice when I open my bag and find a hot laptop with little battery left.

Oh and burn marks in the screen. Lenovo has already replaced one screen.

I have a Dell craptop from the dayjob that sometimes puts itself into hell when going to sleep while unplugged. It gets REALLY hot, probably burning 100% cpu on whatever little nightmare Windows is having, until the battery runs out. If I notice it, pop the lid, and press the power button, it goes right to sleep properly.

To the point of a comment above, this is why the article is noticed when a MacBook is involved, but it wouldn't be if it was, say, a Dell. We just assume that this obviously shoddy piece of generic Windows crap could just catch fire due to design flaws. If my MacBook caught fire, I'd be shocked. If my Samsung Chromebook caught fire, I wouldn't be. If the dayjob Dell caught fire, I'd call it par for the course.

My Macbook does the same thing. Annoyingly frequently I remove it from my bag to find a dead battery and a hot laptop. Worryingly it seems when it's in this weird state, the fans don't run at all.

That doesn't matter; most PCs and Macs (which are just PCs really) will flood the motherboard with power to wake everything up.

Where is that power going to come from if not the battery? you can't have any logic without power.

It ought to have a hard cutoff right at the battery involving only simple electronics, no software, which prevents it from providing the rest of the computer with any power if the battery is at an unsafe temperature.

You can have a small battery or capacitor that powers a small circuit that performs overheating checks pre-boot.

It would be interesting to find out whether Apple's safety mechanism is part of the OS or the firmware.

The firmware on my PC will detect that it is running too hot, and halt. Presumably it will halt to catch fire since it seems as though the only thermal management at that point are the fans, but that is another story.