Preventing Read Disturb Errors in Hot/Cold Storage

The explosion of data generation and consumption has surely ushered in the Zettabyte Era, but in these days of ever-increasing data, storage capacity is not the only crucial consideration when selecting a storage media.?

Your Data Temperature: Searing Hot or Freezing Cold??

There are two distinct categories of data storage depending on its frequency of access: HOT and COLD.??

The origin of these terms may be historically traced to the temperature of their actual physical locations. Data that is retrieved frequently is referred to as "hot data." To facilitate their quick access, they were kept close to hot spinning hard disk drives (HDDs) and central processing units (CPUs). Rarely accessed data is referred to as “cold data.” As such, they were stored on drives or tapes that were archived on shelves away from the warmer data centers. (Source: Backblaze )?

Hot storage describes easy and fast access to frequently used, mission-critical data that cannot wait when you need it. Time is of the essence, so storage for hot data must have low latency, high reliability, and high endurance to withstand repeated operations and to maintain data integrity over extended time periods. Applications like navigation, ecommerce platforms, advanced driver assistance systems (ADAS), and bootup/OS image require quick response times and real-time data retrieval/access and processing.?

Cold storage refers to data that is not immediately needed and is rarely accessed. For record-keeping purposes, they are usually archival, backup, or inactive data. High speeds are not necessary for cold data storage, but excellent retention is.?

Do Not Distub! The Read Disturb Challenge is Real?

A phenomenon known as Read Disturb can jeopardize the integrity of your data regardless of how hot or cold it is.??

Read Disturb causes high read error rates leading to data corruption and data loss.?

How do read disturb errors occur??

Each time a page on the NAND flash is read, voltage is applied not only to the target cell but also to nearby unread cells on the same block, which results in read disturbance from numerous read operations in nearby pages.??

The illustration below shows how read disturbance results in electrical interference and causes the unread cells' threshold voltages to change to logical states other than what was originally programmed.??

After the NAND flash accumulates 100,000 read cycles, uncorrectable errors may occur in the affected pages, resulting to read errors and data failure in the same block.?

Read operations typically account for more than 60% of usage. In read-intensive applications, such as for a bootable device or for a navigation system, the frequency of read tasks are higher.?

Usually, the error correcting code (ECC) could correct the bit errors and restore the data back to its normal state. As error bits accumulate over time, these become uncorrectable even with the ECC engine. When errors reach the ECC capability threshold, the controller would then try to move the data to another healthy block, but by then, the data would already be corrupted and beyond repair.?

Refresh It! ATP’s Data Refresh Technologies?

Data integrity refers to the reliability and accuracy of data over the entire usage life of the storage device. To ensure that there is no loss in data quality, ATP’s e.MMC MLC solutions as well as other NAND flash storage products employ built-in technologies that effectively manage the Read Disturb phenomenon and guarantee the integrity of both hot and cold data.?

• For Hot Zone Disturbance: AutoRefresh Technology?

AutoRefresh Technology improves the data integrity of read-only areas by monitoring the error bit level and read counts in every read operation. It detects when the read count is about to exceed the threshold and copies the data in the affected block to a healthy block before the limit is reached. This prevents the controller from reading blocks with too many error bits and averting uncorrectable data damage. After the re-programming operation is complete, the controller reads and compares the data to ensure its integrity.?

• For Cold Zone Disturbance: Dynamic Data Refresh Technology??

While AutoRefresh focuses on intensive-read areas, Dynamic Data Refresh reduces read disturb and sustains integrity in seldom-accessed areas.??

Data Refresh is used to address data retention and data loss issues that arise from long-term storage in seldom-accessed areas.?

Without affecting front-end read/write performance, Dynamic Data Refresh runs automatically in the background. Bit by bit, it sequentially scans for “cold” areas with flag records. If the number of error bits or read count exceeds the threshold, the data is moved to healthy blocks to prevent the risk of data loss.??

The following figure demonstrates how ATP’s AutoRefresh and Dynamic Data Refresh Technologies prevent Read Disturb errors in automotive navigation applications.?

We are Here for You: WE BUILD WITH YOU

Choosing the right storage for your applications is critical. ATP Electronics is committed to meeting your specialized memory and storage needs with specialized solutions.

For more information on ATP’s AutoRefresh and Dynamic Data Refresh Technologies, e.MMC products offerings, and other ATP Technologies, please visit the ATP website or contact an ATP Representative .