u need to run prime for 24 hours to be verified as stable in benchmarking, some cpu are stable 23 hours and 30 minutes and just hard crash.
2t timing is about 3% difference so this nonesence people come with amd gets hit hard is false
>> u need to run prime for 24 hours to be verified as stable in benchmarking, some cpu are stable 23 hours and 30 minutes and just hard crash.
#1 Stability is all relative. A system might not pass the so called "cpu stressors" for more than a few hours yet may never have a stability related crash in normal usage. It all depends on what your using the system for. That said its a good starting point.
#2 Prime isn't perfect. Passing Prime for x hrs doesn't mean 100% stable nor vice versa.
#3 The reason why 24hrs is picked...is because it represents a full day cycle. It is thought that only a full day testing is enough to test all possible normal environment variations. Which of course is also imperfect since there are more environmental (and other) variables to consider. A single unusually hot or cold day can skew results for example.
>> 2t timing is about 3% difference so this nonesence people come with amd gets hit hard is false
If your benching mem bw intensive appz then it matters more than the "3%". If not, then it can impact everything from 0% to that "3%".
The point is, these values only really matters if your trying to push the limits. Which is what the people in here are trying to do no?
Therefore that "3%" does matter in this context. Otherwise you not get high performance ram in the first place. Most hardware upgrades are improvements (if at all) in small percentages like this. So its almost as important as say a cpu or motherboard upgrade in that context no?
"Performance is what makes MHz useful" (tm) Synth.
In general more agressive timings will net you greater overal performance vs raw Mhz.
By itself the % may not seem much, but when part of the sum of performance tweaks...it adds up and matters.
Show Posts
