July 20, 2026
Bit flips and comment meltdowns
ECC and DDR5
Your memory may be fixing its own mistakes — and the comments are losing it
TLDR: The article says newer DDR5 memory has a built-in error fixer, but it still doesn’t replace full protection that can warn you when something is going wrong. In the comments, people split between wanting better safety, demanding actual evidence, and fuming over wild price swings.
The big idea in Russell Coker’s post is simple: some computer memory can catch and fix tiny mistakes before they wreck your files, and he thinks that protection should be far more common. He walks through why traditional error-checking memory matters, why newer DDR5 memory has a limited built-in safety net, and why that built-in feature is not the same as full protection. But the real fireworks were in the comments, where the crowd turned a nerdy memory post into a full-on consumer rights, pricing rage, and "show me the receipts" showdown.
The spiciest reaction? A huge side-eye at the suggestion that governments should push people toward safer memory. One commenter basically said: slow down, that’s a massive overreach. Another camp was less interested in mandates and more annoyed that if DDR5 memory quietly fixes errors inside itself, users often can’t actually see the error count, which means your computer may be whispering “I saved you” and never telling you it almost messed up. That sparked the classic tech-forum split: people with real-world horror stories versus data skeptics demanding proof. One user said their memory catches errors every couple months and that without protection they’d never even know. Another fired back with the forum equivalent of pics or it didn’t happen, arguing everyone keeps repeating warnings without fresh numbers.
And then came the most relatable comment of all: the price drama. One user claimed 16GB error-correcting DDR5 jumped from $50 to $500, then joked that when the AI bubble pops, bargain hunters will feast. In other words: part safety debate, part transparency fight, part waiting-for-the-crash meme.
Key Points
- •The article explains that ECC RAM commonly uses Hamming codes to correct single-bit errors and detect double-bit errors before corrupted data reaches the CPU.
- •It distinguishes RDIMMs from UDIMMs, describing RDIMMs as buffered modules that support larger memory capacities and more DIMMs at the cost of slightly higher latency.
- •The article says ECC UDIMMs are often expensive on the second-hand market, while ECC RDIMMs can be relatively cheap because retired servers release large quantities of them.
- •It describes ChipKill as an advanced ECC approach designed to tolerate failure of a single DRAM chip, and notes that similar features are marketed by Dell and HP as Advanced ECC.
- •The article states that DDR5 includes on-die ECC to mitigate higher error rates from denser, faster memory, but says this is not a replacement for conventional motherboard or system-level ECC.