By the way, the reason the address->IC mapping is non-trivial is for performance reasons.

If you were, for example, iterating over a column of a 2D array with a stride equal to some power of two, you'd end up hitting the same IC repeatedly. This limits your bandwidth to that of a single IC, rather than balancing it over all of them. (I'm simplifying a little, but that's the gist)

So they try to "swizzle" things, using a formula like the one in the article.

Not loading for me, but here's the archive: https://web.archive.org/web/20250207044221/https://blog.pier...

On 80's arcade games and home computers you can try piggybacking a known working RAM chip onto each of the RAMs for a quick diagnosis, if you're lucky (it doesn't always work).

If you stress test small address ranges around each failing address, does that selectively heat up the chip to which that address routes?

Then all you'd need is an infrared imager.

Alternatively, could you stress test the failing addresses while selectively heating each chip in turn with a hot air gun and checking if the error rate changes?

Is the normal person going to de-solder this? Come on!!!

For HW hackers playing master it is ok. But for the average person it is not an option.

You can't change your RAM and this is disastrous.

Even normal computer shops may torture you for weeks (after you have been torured by the mother board)

This is Hacker News. You can do IC swaps like this with <$100 of equipment. Not doing so is a choice.

Ok, but how much do the skills cost?

noob

Soldered RAM is pure evil.