ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
Remaining unbiased and consistent are my top priorities. I do not have any personal stake in ncase, Noctua, asus etc. The results are the results whether I like them or not, and whether they make sense or not.
2
3
I use quite a few tools to make the results as consistent as possible. Things like a window ac unit so I can control the temperature of the room, thermometer away from the ac airflow and case path, decibel meter, hwinfo64, google sheets, argus monitor, msi afterburner and Intel xtu. My testing is done either in the morning to early afternoon or at night. During afternoon hours the sun starts pouring in my window and affects the room temperature.
4
5
The programs I use to test are heaven (latest version) and prime95 v26.6 (new C14S testing uses 30.3 no AVX) small fft's and they are run simultaneously – although I am looking for an alternative GPU stress test.
6
7
First step is to tune the system and fan curves, dial in the overclocks, make sure everything is stable.
8
9
Second is to control ambient at 20c.
10
11
Next, I will load only the necessary programs for the test (google sheets, heaven, p95, timer, hwinfo64) and kill background applications like malwarebytes, discord etc.
12
13
Before test one I will run a trial for ten minutes to saturate the sinks in the system and check that hwinfo64 is showing the correct overclocks, voltages, wattage etc.
14
15
I let the system idle for 20 minutes and then I can begin testing.
16
17
I start with typing in my ambient and begin values in google sheets. Then I begin initializing heaven since it takes a few seconds to load. As soon as heaven starts running, I start the p95 test, then reset hwinfo64 temps and have it record the information in a file, then I hit the timer that runs for 20 minutes.
18
19
Once 20 minutes is up, I will stop the ac unit to record noise levels, then record ambient and stop the hwinfo64 logging and stop the stress tests. Then I take the data from the log and put it in the spreadsheets. If there are any outliers, I will throw out the data and run the tests again. If there are any amazing results, I will restart the computer and test again. I need to be able to duplicate the results within a 3c range (my arbitrary number) to verify the result.
20
21
Each successful test from start to finish takes about an hour if you calculate warm up and cool down time. Once I am satisfied that a result is good and repeatable then I will adjust the fans to a new configuration and start over again.
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100