Microprocessor-based devices may experience crashes due to failures in software, hardware, or a combination thereof. Details about a crash may provide valuable information for determining cause of the crash. Devices may save crash information in volatile memory, place the volatile memory into a self-refresh mode, power cycle the remainder of the device, reboot, and retrieve the crash information. Other information to be persisted across a power cycle may also be saved to volatile memory in this fashion as well.