My 7 is on the December one.

Aside from the cost, ACQ107 is not very reliable in my experience. I have one that randomly drops the connection every now and then, even at sub-10Gbps speeds. Switching to a different NIC makes everything rock stable.

In my experience (Im running several of these) thats down to heat.

I redid the thermal paste on my cards and haven't had any issues since.

I've got several, they've been fine for me. That said modern NIC variants in general seem to be a crapshoot, regardless of speed/price.

In my experience every single (AIO) water cooler I've tried has been significantly worse in terms of noise than a decent air cooler.

Specifically the pumps. Those things have an obnoxious high-pitched whine that I personally find unbearable, especially during low/idle workloads.

It's possible that the actal dB level is lower, but the frequency and sound characteristics matters. A lot.

> libc is big and complicated and most programs only use a tiny fraction of it. Is libc a problem for the ecosystem?

IMO yes. I definitely believe having basic common functionality (malloc, printf, memcpy etc.) provided by one library with all the crazy/obscure stuff that very few people need or want somewhere else would be an improvement.

You might want to check the date.

Because that would break compatibility. The way shadow stacks are implemented means they can be enabled in existing software without code changes.

If one were to design a modern ISA from scratch it would make sense though.

> event 1 is always “Site wiring” and every other event is always “None”, even though there have been multiple power events, and there is no site wiring issue

Some UPS:es are adamant about wanting live and neutral on specific pins on the power plug and will throw that error message if they're swapped.

If it's a reversible power plug, flip it around. If the power plug can only be connected in one orientation the outlet is likely incorrectly installed (which is not all that uncommon, as 99.99% of stuff will work perfectly fine with live and neutral swapped).

Yes, for 32- and 64-bit registers. Most modern x86 CPUs has fast paths for 'xor reg, reg' which performs the zeroing using the register renaming mechanism instead of actually executing anything on the back-end. So the only cost is that of decoding the instruction.

Recent Atom CPUs (since Gracemont) do support AVX and AVX2.

The point is that avoiding Apple and Google, while it may be technically and theoretically possible, is going to make your life extraordinarily complicated and is completely unrealistic for the average person in our modern society.