Hardware
ECC Memory
Also known as: Error-Correcting Code Memory, ECC RAM
ECC (Error-Correcting Code) memory detects and corrects single-bit memory errors in real time, preventing data corruption that could crash a server or silently corrupt data. Required in all production servers.
Standard RAM (non-ECC) has no mechanism to detect memory errors. A cosmic ray or electrical noise can flip a bit — in a workstation this might cause an occasional crash, but in a server running a database or hypervisor it can cause silent data corruption without any indication something went wrong.
ECC memory adds extra bits that let the system check every read for errors — silently correcting single-bit errors on the fly and detecting larger ones before they cause damage.
ECC is a requirement, not an option, for any production server: database servers, file servers, hypervisor hosts, NAS systems. The cost premium over non-ECC is small and the risk of running production workloads without it is not worth taking.
Consumer processors and motherboards (Intel Core, AMD Ryzen) often don't support ECC even when the memory module supports it — ECC requires explicit support from the CPU and memory controller. Server platforms (Intel Xeon, AMD EPYC) universally support it.