GeIL (that's capital 'I' capital 'L') is going Hollywood with its naming scheme for a new technology the company claims will result in higher quality memory shipping from the factory. Called D ie-hard B urn-in T echnology (DBT) , GeIL says the new system will virtually eliminate early failure among memory modules and catch defects that otherwise would have went unnoticed.
This begs the question, don't memory manufacturers already test for defects to avoid this scenario? Yes, and here's what GeIL has to say on the matter:
"The conventional way of memory module burn-in is done by regular market-available motherboards with limited module accommodation and burn-in time of approximately 20 minutes under room temperature. Not to mention that each of the motherboard-based burn-in platform requires manpowered monitoring." - GeIL
Using a custom-made burn-in chamber, which the company calls DBT-1, GeIL claims it can test up to 1,000 modules all at once. Instead of using traditional motherboards, the DIMMs are placed onto specially designed chipsets outfitted with GeIL's custom coded burn-in testing software. Temperatures are then ramped up to 100 degrees Celcius where the modules sit and roast for up to 24 hours, a process GeIL says ages the RAM to the equivalent of three months of normal usage, or just beyond the time that most early failures are likely to occur.
According to the press release , all GeIL products will eventually be subjected to the rigorous DBT burn-in, but not right way. Kicking off the new technology will be select DDR-2 kits, with notebook SO-DIMMs and other memory soon to follow. All products certified with the DBT process will carry a special logo, seen below.
Image Credit: GeIL
Maybe, or maybe not. Cutting through the marketing hype, GeIL brings up a point that not only applies to system memory, but many PC components. Hardware failures, if they're going to occur, typically take place sooner rather than later. If your videocard has a bunk GPU or your new memory sports defective chips, you can expect quirky behavior right from the get-go.
As it pertains to memory, gone are the days when you had to worry about getting hold of C-grade stock , or chips deemed to have defects. Generic manufacturers would snag these partially defective chips for pennies on the dollar, slap them on a PCB, and throw them into the marketplace. Today, lifetime warranties are the norm, not the exception, and you won't find many oddball or otherwise generic manufacturers when buying from a reputable vendor.
Just because you're not equipped with a custom built torture chamber with 1000 degree thermals doesn't mean you can't uncover dormant defects waiting to rear their ugly heads. The best way to ensure your RAM is working is to simply use it. If your system blue screens, freezes, refuses to boot, or displays other quirky symptoms out of the ordinary, you might have a bad kit. To test for faulty DIMM, or for some peace of mind, download and run the free Memtest86+ utility. You can install Memtest86+ onto a CD, DVD, floppy disk, or USB key, and then configure your BIOS to boot directly from whichever media you chose. Memtest86+ will proceed to stress test your RAM, and while it can take up to hours for defects to appear, most times faulty kits will be exposed right away with anywhere from hundreds to thousands of reported errors.
As with most PC components, too much voltage can be a surefire path to premature death. Your BIOS should auto-detect both the correct voltage and latency settings for your RAM, which come pre-programmed onto a special section of your RAM, but the BIOS doesn't always get it right. When installing a new kit, double-check that you're running the correct voltage as specified by the manufacturer, and peek at the latency settings too. Running overly aggressive latencies isn't likely to kill the RAM, but it can set off a false-positive by churning out errors in Memtest86+ and displaying other symptoms of a bad kit.
Image Credit: Wikipedia