Kernel version： QT 4.1.15 and 3.10.17
CPU：i.MX6Quad
I had a problem with the new motherboard, which crashed while running the official GPU test script GPU. SH. It ran at room temperature, and the runtime was irregular. Sometimes it was ten minutes, sometimes it was a dozen hours。
Measure several CPUs and voltages while running the GPU test script, and test several voltage waveforms when the system crashes：
Question 1：
How does IMX6Q's power VDDSOC_CAP/VDDARM_CAP/VDDPU_CAP run so fast when running GPU test? Is this normal?
Comparing our PCB and official differences, we found that there are differences in voltage filter capacitors under CPU。The main control chip supply voltage VDDARM_CAP, VDDCORE, DDR_1V5, VDDSOC_IN, VDDSOC_CAP and VDDPU_CAP and the development board are different, our board filter capacitor position is as follows:：
Official development board master chip power supply voltage VDDARM_CAP, VDDCORE, DDR_1V5, VDDSOC_IN, VDDSOC_CAP and VDDPU_CAP filter capacitor position as follows:
Question 2：
The difference between the location of the filter capacitor and the development board is different. Will this difference have any effect? What's the impact? If the capacitance is affected, how can I increase the capacitance on our board? Also, the 22uF filter capacitor of the development board is removed, and it will not crash. Is PCB wiring affected?
