427 karma · joined November 10, 2025
I've tried to find this graphic against several times over the years but it's either been scrubbed from the internet or I just can't remember enough details to find it. Amusingly, it only just occurred to me that maybe I should ask ChatGPT to help me find it.
And then there are other factors like more sticks of ram stressing things further. I had to downclock to get memtest stable when running 4 sticks even though each kit ran fine on it's own. But that is expected as well as 4 sticks stresses the memory controller even further.
I confess I don't have any real recent experience with DDR5 though, mostly with DDR4 on Ryzen 1000-5000 series.
Meanwhile you've got OpenAI buying up all the DRAM and maybe just piling it in a warehouse so no one else can have it or they figure out what to do with it. Microsoft recently said they don't even have a power source to plug the ton of GPUs they just bought into so they're also just sitting around collecting dust. What is even happening?
There's a lot of blame for parents, much of it deserved. But when you have CPS being called for kids playing in the woods or parents charged with manslaughter when some one else runs over their kid you realize this is now going against the grain to resist this stuff.
I never used it really myself. The original UI wasn't what I'd ever want out of a PC but the impending stench of the Windows Store was what drove me off of Windows at that point.
I have an 8.1 VM in my unraid server that only exists to run an older radeon driver that allows the GPU to turn off to near 0 watts idle when the hardware isn't in use. Windows 10 broke the subsystem that these drivers used and AMD never got this feature working on 10.
I've seen this topic come up a lot on a datahoarder forum but there was never any consensus on what it would take to assure the data was refreshed because no one really knew how it worked. People would assume anything from a short power up to simply rewriting all data as possibilities.
The firmware engineer responding up thread is actually the best information I've seen on this. And it kind of confirmed my suspicions. He works on enterprise drives and they run a refresh cycle as background cycle that seems to start on a random timer. I didn't see any response (yet) on how long this process takes but I think we can infer that it takes a while based on it being a background process. Probably hours or more. And this is all for an enterprise drive, I wouldn't be surprised if consumer drives have a less robust or even non-existent refresh function.
I'm of the opinion that that based on the little information about some implementations of this function, the logical conclusion is you should just rewrite all of the data over again on your cold backups for the cycle. Powering it on isn't enough, even powering it on for a day might not do anything. As you said this is a pretty big drawback for uses flash as a cold backup.