Embedded memories are the most dense components within a system-on-chip (SoC), accounting for more than 50 percent of the chip area. Designed using aggressive design rules, embedded memories tend to be more prone to manufacturing defects and field reliability problems than any other core on the chip. Therefore, the overall yield and reliability of an SoC depends heavily on the memory yield and reliability. Memory failures can be attributed to manufacturing defects and transient or random soft errors. For self-healing designs, failures due to both manufacturing defects and transient errors need to be addressed.
Manufacturing defects, as the name suggests, are permanent in nature and are introduced during fabrication. Manufacturing defects affect the SoC yield and need to be mitigated to achieve lower silicon cost. Although the yield of native memory may be inadequate, embedded memory yield can be improved through the use of redundancy or spare elements. An effective solution to heal or repair these defects must not only determine the type and quantity of redundant elements needed for a given memory, but must also be able to detect and locate the defects in the memory while understanding the allocation of the available redundant elements to maximize repair. Synopsys’ solution allows in-system power on BIST and repair to automatically detect and heal manufacturing defects in the field.
In addition to manufacturing defects, memories are prone to transient or random soft errors that can lead to data corruption or cause electronic systems to crash. These random errors are functional errors resulting from strikes by energetic ions such as neutrons and alpha particles. Transient errors can be catastrophic and need to be mitigated especially for the mission-critical applications, including networking, high-end computing, and space applications, which require high availability and reliability. Systems need to allow data being accessed from memory to be checked for errors and corrected on the fly, enabling the memory to self-heal.