I'm considering buying two 16GB DDR4 sticks second hand. The seller says they came from an Acer Nitro V laptop and can install them in another laptop to demonstrate that they work. The problem is that a full MemTest86 run could take around three hours, which is inconvenient given the travel distance. I have used OCCT's memory test before and it would take roughly two hours for both sticks. Is a clean OCCT result sufficient for evaluating used RAM, or is MemTest86 still necessary?
3 Answers
OCCT and MemTest86 aren't exactly doing the same job. OCCT is primarily a stress and stability test, while MemTest86 is more focused on diagnosing memory errors. Ideally you'd use both, although longer tools such as Y-Cruncher or TestMem5 can take several hours too. For a quick second-hand purchase check, OCCT is useful, but it isn't a perfect replacement for a thorough diagnostic test.
OCCT's newer memory test is a reasonable practical check, especially if you're trying to avoid spending several hours on-site. A clean result would make me fairly comfortable, but remember that a single error is enough to reject the RAM. The seller might be tempted to keep rerunning the test until it passes, so don't treat repeated failures as something that can be ignored.
MemTest86 is still the more traditional choice for checking RAM, while OCCT is also commonly used for processor and graphics stability testing. If time is the main concern, OCCT can provide a useful screening result, but a short test cannot guarantee the sticks will never produce errors under every workload.
That's the part I'm weighing up. MemTest86 is appealing, but spending several hours there just to test two sticks is difficult.
That last point is really helpful. I hadn't considered that someone might keep rerunning the test until they got a clean result.