Benefitting From ClearNAND Technology

In the beginning there was raw NAND.* While NAND was a giant leap forward in providing high storage density at a more reasonable cost-per-bit than NOR, it came with some inherent flaws that had to be managed. In addition to traditional block and page remapping requirements, factory bad blocks and bit errors that eventually lead to block failures required the addition of complicated (and expensive) controllers. As lithographies shrank largely to control costs, a lower erase count ceiling magnified the impact of inefficient wear leveling and exacerbated NAND’s inherent flaws, requiring even more complex (and expensive) controllers; a vicious circle of cost pressures and endurance/performance issues was born. NAND has now reached the point where these external controllers are challenged to keep up with the error correction required by the tiny sub-20 nanometer chips, and as a result exact a significant performance penalty. 

Cost pressures lead to shrinkage of NAND cells which require ever greater ECC/EDC

 

*Actually in the beginning there was EEPROM and then NOR, but that’s a story for another time.

Memory manufacturers have solved some of these issues with managed NAND, a method of packaging raw NAND with a hardware-based controller including error-correction (ECC), wear-leveling, and bad block management. For OEMs, integration of managed NAND (eMMC, for example) is much simpler and the resulting product is more reliable than raw NAND. The price however, comes in the form of a performance penalty due to the inefficiency of running these NAND management functions with a non-die controller, which must be inexpensive to control cost, and thus of limited capability.

Comparison table of ClearNAND/EZ NAND to other NAND-based storage options

ClearNAND was developed to solve this problem. On a ClearNAND chip ECCs are performed on the die, and have the depth to match the NAND requirements. This leaves other more complex NAND management functions, such as page and block remapping, wear-leveling and bad block management, to be done in software, utilizing the much more powerful system processor and giving OEMs the best performance available while still maintaining access to control how the NAND is managed.

The combination of Datalight FlashFX Tera and ClearNAND, for example, provide an ideal solution for device manufacturers who want to take advantage of this new technology. With support for over 300 flash parts “out of the box,” FlashFX Tera also enables OEMs to upgrade flash parts easily, making them more competitive and speeding time to market.

To learn more about FlashFX Tera’s advanced wear-leveling and bad block management, visit our product page.