AMD Test System and Temperature Results
AMD Test System |
---|
Processor: | AMD Ryzen 9 3900X 3.85 GHz Base / 4.0 GHz OC |
---|
Motherboard: | ASUS ROG STRIX B550-E GAMING |
---|
Memory: | 2x 8 GB Crucial DDR4 3200 CL22 Model : CT2K8G4DFRA32A |
---|
Graphics Card: | MSI GeForce GTX 1650 Ventus XS OC Edition |
---|
Storage: | Crucial P2 500 GB M.2 NVMe SSD |
---|
Power Supply: | Chieftec Chieftronic PowerPlay 850 |
---|
Case: | Thermaltake Core P3 |
---|
Operating System: | Windows 10 Professional 64-bit Version 20H2 (October 2020 Update) |
---|
Thermal Paste: | Arctic MX-5 |
---|
Other: | Crucial Ballistix 2x 4GB DDR4 2400 Model: BL2K4G24C16U4B (clearance testing only) |
---|
Testing Procedure
All testing is done at a room temperature of 22°C (72°F) with a 1°C margin of error. For the AMD test bench, the CPU's 3.85 GHz base all-core clock is obtained using the AMD Wraith Prism cooler with the fan at maximum RPM. Once the all-core clock is verified, we limit the CPU to this frequency under load for the stock tests. That said, all other features are enabled, allowing the CPU to clock down at idle. All tests are run with the CPU cooler's fan pegged at 100% PWM for the base results. Meanwhile, all coolers are tested at a noise-normalized 45 dBA to demonstrate their performance at an acceptable noise output for the majority of users. As for the overclocked results, the AMD Ryzen 9 3900X has been pushed to 4.0 GHz at 1.22 volts.
Refer to this articlefor a full breakdown of our testing procedures.
The idle test consists of the CPU sitting idle at the desktop for 15 minutes.
The primary load test consists of the Blender BMW render looped for 15 minutes, and the peak temperature is used as the result.
In the final test, we use AIDA64's FPU stability test, which ensures maximum heat generation. While it is typically considered an unrealistic workload, it is used to show the overhead a CPU cooler has available in a worst-case scenario.
All tests are completed three times and averaged, for both maximum performance and noise-normalized results.
Idle Temperatures
Typical Load Temperatures
Maximum Load Temperatures