The testing workflow behind trustworthy server-pull memory, what SPD data tells you, and why a serial-number list protects the buyer.
Published August 2026 · AltoChip team
Pre-owned server memory is a excellent-value category with one catch: you cannot tell a good module from a dead one by looking at it. The entire difference between a bargain and a problem is the testing and documentation behind the lot. That is why serious buyers evaluate the seller’s evidence process before they evaluate the price.
A dependable workflow runs every module through fixed steps: visual inspection of contacts and PCB; verification of part number, capacity, speed grade and rank against the lot list; a memory test pass (MemTest86 or equivalent); and Server POST or live-system validation on a representative platform. Failing modules are removed from the lot, and the lot is recounted against the serial-number list before packing. This is the workflow AltoChip runs — described in full on our Testing & Evidence page.
Every DIMM carries an SPD (Serial Presence Detect) EEPROM holding its manufacturer, part number, speed bin and timing data. Reading SPD confirms that the label matches the silicon — a basic but essential check against relabelled or misdescribed modules.
A per-module serial-number list ties the physical goods to the paperwork. It lets you verify on receipt that what arrived is what was tested; it makes any DOA claim precise (this SN, this symptom); and it prevents disputes about substituted modules. When buying pre-owned memory from any seller, ask three questions: is there a per-module test result, is there an SN list, and will you see photos before paying? If any answer is no, price the risk accordingly.