Researchers develop algorithm to distinct between recycled and new NAND flash chips

Researchers of the University of Alabama have developed an algorithm that can distinct between new and recycled NAND flash chips and report the age of the chip. The technology should protect users against shady recyclers that sell used NAND flash chips as new. A smartphone could run an app with the algorithm.

The technique can distinct between used and new chips, even if the chips have only been used 3% of their normal lifetime. Currently, NAND flash chips are recycled without durability tests. If the chip is functional, it will pass most tests. With the new algorithm it's possible to see how much the chip has already been used.

The method works by making use of the fact that the more a flash cell is programmed or erased, the more defects will occur. The defects can be measured because the more defects, the more current leaks when the flash cell is supposed to be off.

"It also reduces the rate at which charge moves through the device. These effects show up as a slowdown in the memory’s erase time", according to the researchers who developed the algorithm. By measuring how much the more time it costs to erase a cell, the age of the chip can be determined.

The researchers are able to determine the age of a flash chip with a confidence of 100%, even if the chip has been used less than 3% of its lifetime. But that's not enough for them, a new method that is in the works should even better estimate the age of chips. This method works by measuring how often bits will flip after the chip is hammered.

"It turns out that if you “hammer” a page (about 16 kilobytes of memory) by programming it over and over again, a certain number of bits will flip. Data that should be zeros will be ones, and vice versa. Older flash, even if it’s only gone through 1 percent of its lifetime, will reliably have more of these flipped bits than new flash will," the researchers explain.

While the erase method can work on a smartphone, e.g. determine the age of the flash memory of the phone, the hammer method might not. According to the researchers that method will "will take some work".

No posts to display