The ASRock B550 PG Riptide uses a single, large heatsink for the CPU VRM section. Due to its size, heat absorption takes longer, but without a set of small fins, heat will also linger for a longer period of time. No heatsink is present for the SoC section, which can be a concern for APUs that primarily use the SoC voltage rail for the iGPU.
Ryzen 9 3900X Stock CPU
CPU Voltage:
1.025–1.494 V
DRAM Voltage:
1.35 V
Idle Power:
19 W
Load Power:
130 W
Ryzen 9 3900X 3.8 GHz CPU
CPU Voltage:
1.4 V
DRAM Voltage:
1.35 V
Idle Power:
25 W
Load Power:
211 W
VRM Voltage Probe
Ryzen 9 3900X Voltage Check
Load-Line Calibration
Coil Readout
SV12 Readout
CPU LLC AUTO
1.468 V
1.325 V
CPU LLC 1
1.488 V
1.340 V
CPU LLC 2
1.468 V
1.325 V
CPU LLC 3
1.456 V
1.131 V
CPU LLC 4
1.438 V
1.300 V
SoC LLC AUTO
1.220 V
1.140 V
SoC LLC 1
1.220 V
1.156 V
SoC LLC 2
1.220 V
1.150 V
SoC LLC 3
1.200 V
1.150 V
SoC LLC 4
1.198 V
1.150 V
Using a multimeter probe, I measured the voltage on the back of the coil under load. This measuring point is after the 12 V step-down occurs. With a fixed voltage of 1.4, I took the measurements during load and proceeded with this process for each LCC the BIOS offers for selection. A second readout has been taken from the software readout, listed as SV12. This value is measured by the CPU as the current voltage input.
For temperature measurement, I use a Reed SD-947 4 channel Data Logging Thermometer paired with four Omega Engineering SA1 Self Adhesive Thermocouple probes. One probe directly touches the chipset and two are placed on select power stages. The last probe actively logs the ambient temperature.
For the ASRock X570S PG Riptide, one probe is placed along each bank of power stages. A probe is left out to log the ambient temperature. All temperatures are presented as Delta-T normalized to 20 °C, which is the measured temperature minus the ambient temperature plus 20 °C. The end result accounts for variation in ambient temperature, including changes over the course of a test, while presenting the data as if the ambient were a steady 20 °C for easy presentation. Additionally, there is direct airflow over the VRM for the first five minutes, after which the fan is removed. This gives an idea of what to expect with and without moderate case airflow.
Prime95's Small FFT is used for maximum power consumption over a 30 minute period. For testing, I used a Ryzen 9 3900X set to 3.8 GHz and locked at 1.4 V and 1.15 V for the SoC. This allows the benchmark to run longer and avoids overheating the CPU. A common misconception is that the maximum clock frequency inherently has higher power draw. While this certainly can be true most of the time, it also comes with an abundance of generated heat. Therefore, a lower clock speed with a higher voltage actually draws more current in this instance, which allows me to run tests longer without fear of burning out the CPU as it approaches 90°C from a long sustained power load. Temperatures are logged every second, and the two probes are then averaged for a cleaner presentation before subtracting the ambient to calculate the Delta-T. The results are charted below.
The ASRock B550 PG Riptide did poorly in the VRM torture test. With a fan directly on the VRM section, the SoC tapers off around 35 °C, and the CPU phases in the low 40s. Once the fan was removed after 5 minutes, temperatures quickly rose until the SoC reached 60 °C, only increasing from there on out as the room temperature increased. The CPU portion increased until reaching 86 °C before the test ended. I want to point out that the data sheets for these Vishay SIC654 power stages do not specify any thermal parameters. I therefore cannot comment on the longevity of said component at these operating temperatures under a forced, stressed load scenario.
One issue I didn't think about until after I finished all the testing was the affect of using the motherboard's default Auto LLC. In future torture tests, I will make sure to lock and report this value to better represent a consistent, repeatable test. When compared to the X570 PG Riptide VRM tests, temperatures for the CPU section are higher even though both share the same components and applied voltage. This can only be explained by a different AUTO LCC applied by the BIOS.