Tried to update via F7 menu on boot to 0051. Froze halfway through. I left it for about a half hour before it shut off. Now when I try to turn it on, the LED turns off IMMEDIATELY. Holding power to access the POWER menu does nothing. Removing the jumper does nothing either. This is my second NUC that has seemingly bricked. Out of the two, they've only worked a combined hour before crashing and burning. Removing RAM doesn't yield the 3 beep/flash code.
Contacted INTEL support but hoping there's a troubleshooting step I've missed.
Nope, no missed steps. You have to wonder whether there is something you are doing that is responsible for this. As an example, bad or improperly installed SODIMMs or M.2 module could be responsible for damaging the system (creating a short, etc.).
I suggest that, while you wait for the new NUC to arrive, you get these SODIMMs replaced as well. When you do get the new unit, test thoroughly and update the BIOS before installing any storage. For example, download the ISO for MemTest86+ and install it onto a flash stick. When you power up the system for the first time, insert this flash stick and test the memory (I recommend for 24 hours, so it is also a burn-in for the unit).
Well both NUCS had 2 different and both brand new sticks of RAM for a reputable company in them when I started them and both ran mostly fine (graphics got choppy until BSOD which I've read was because of the dated BIOS) until I tried updating the BIOS. I think it's equally unlikely for 2 units of RAM to be messed up.
Memtest is new to me but i have nothing to run it on that takes this type of RAM. Also, you said to test the BIOS on the new unit thoroughly. How would one do that?
No, I (perhaps poorly) said that (a) you should always test your RAM thoroughly (and use this testing as a burn-in test for the system) and (b) upgrade the BIOS before installing any storage. In a perfect world, because of the many compatibility issues that can come up with memory, I would prefer to upgrade the BIOS before testing the RAM, but the dependency chain is problematic.
Your idea of reputable and mine is probably a *lot* different. I am sure that there are companies you consider reputable that I won't touch with a ten foot pole. Suffice it to say, with the sample-based testing that most of the industry is using (like, "oh, let's test one in one hundred; that should give us reasonable coverage" - NOT!), there is a measurable chance of getting a pair of DIMMs that are both bad. As Andy Grove used to say, "Only the paranoid survive"...