Flash Features in Automotive MCU Chips

Unlike consumer and industrial MCUs, the Flash inside automotive MCU chips is generally divided into PFlash (Program Flash) and DFlash (Data Flash). EEPROM is usually simulated through Data Flash. Another difference is that automotive MCUs are basically Embedded Flash, and rarely use XIP (eXecute In Place) method.

1

Parameters of PFlash and DFlash - Endurance, Page Size

The following figure shows the endurance of PFlash and DFlash of TC336LP. Generally, the default DFlash configuration is Single ended sensing mode (the available size of DFlash in Complment mode is halved). In this case, the erase operation of DFlash is more than 46 times that of PFlash.

2

PFlash Parameters - Wait Cycle

A long time ago, when Xiaoer was learning microcontrollers, he found that the program often ran wild. Later, he found that the reason was that the Flash Wait Cycle of the chip was inexplicably modified.

The CPU speed is very high, such as TC336, which can run up to 300MHz, but the speed of PFlash is not enough, and response time is required, otherwise the reading result will be wrong; this is the origin of Wait Cycle;

As shown below

As you can see, local access to PSRR is 0 wait, that is, single-cycle execution. For some time-sensitive programs, they will be put into PSPR to run;

Similarly, as shown in the figure below, data reading and writing are also the same

In this article, I guess everyone has understood why increasing the main frequency alone has limited impact on computing performance.