Industrial & Military standard SSD

Firstly, let’s see the conclusion that Renice draw based on the test from Nand Flash test platform NFA100. 1, the P/E cycle of some Block up to 20,000 times in real test, though it said only 3,000 times in datasheet. 2, Each block is an independent individual, and the initial condition of each block is not exactly the same, even though they are the same process in the same production line. 3, No matter it’s original Nand Flash or non-original, they have the necessity of testing, the Nand Flash can be sorted and classified into different levels. 4, Though it is the same Nand Flash chip, the health condition of each Block is not the same 5, The real lifetime (P/E cycle) of Nand Flash may be far greater than its datasheet, only when test, it can play the greater value. 6, Deep understanding of the life performance of Nand Flash in different P/E stages can help better plan ECC. 7, For those who want to convert MLC mode to SLC at a certain stage, when to start ...

If all Page performance is basically the same when programming every time, this Nand Flash chip should be the most stable and reliable. Therefore, you can know the reliability of Nand Flash by observing the distribution of each Page performance. Although the datasheet for each Nand Flash chip will indicate detailed time of Erase, Program, and Read Performance , in fact, these Performances average the Lower page and the Upper Page, so it will mislead people that the performance of each page is the same, actually it is not. Renice test setting and idea: Take a new 3D Nand Flash for example, choose 1024 (03FF) Block, Program/Erase for 1000 times. The more stable the graph shows, the more reliable the Nand Flash is, if the graph floats drastically, the worse stability of the Nand flash. It looks stable when Page0 program A not ideal Lower Page test result as following. A not ideal Upper Page test result as following. Conclusion: the...