The best comparison for FRAM vs. Flash write speed can be seen when comparing two MCUs writing to non volatile memory at equal clock speed. In this case the flash based device is pitted against a FRAM based device such as the FR59. Both devices are running at a CPU speed of 8 MHz and executing real world application code that does pointer updates, data handling using similar routines. The FRAM device outperforms the flash device by a factor of 100. Flash writes are capped at 13 kBps and this includes the erase time per segment write, whereas an average use-case FRAM write can go up to or even faster than 2 MBps. One example application where such high speeds are game changing is when doing over the air updates or reimaging the firmware on a target device. Writing 64 k of data is no longer confined by the bottleneck due to slow writes/erases of non-volatile memory and the device firmware can be changed 10 s of thousands of time with a frequency that is only limited by the communicating protocol. Think about an application receiving streaming data over a SPI or I2C bus. However, with FRAM there is no need to buffer this data and it can be moved directly from the bus to non-volatile FRAM without any bottle necks due to flash setup and erase. Note that another important feature with FRAM is that since the writes happen within the instruction cycle at 8 MHz, the CPU is not held or wait-stated as in the case with Flash writes.