Originally Posted by Nitethorn
Just a thought from the outside looking in, I think the reason people post "maybe you should give up" is because while yes, this is a community where we try to help each other but a lot of your responses, the way you word yourself and the caps with lots of question marks comes across as very abrasive. People are trying to help you, have been for weeks now. There is no magic "plug this in and it'll work". Without having that kit right in front of us to test, it's impossible to just "make" it work for you. I know you probably don't mean to be so aggressive about it, but I think you should take a deep breath and realize that even the "unhelpful" posts are people that are literally trying to be helpful in some way. Just my 2 cents...
The intent was never to be abrasive, and if you read my message above, specifically to
, I thanked him numerous times (despite not making any progress).
I know people are trying to help, but part of my frustration stems from the "advice" I am receiving, is not actually tailored to my knowledge.
Most of the things people are saying are highly technical, and no one is actually explaining the rationale behind those suggestions, so it's like I am randomly typing in numbers that people are suggesting, without any logic, and praying that something works.
If this community is meant for only helping other overclockers that know what they are doing, then so be it, but I didn't get that impression.
The caps, are to emphasize points.
You can see that part of the frustration is from missing information, unanswered questions, or flat out conflicting information. For example, I have asked the following question a dozen times already, yet not a single person has been able to answer. The question was, why is Ryzen Dram Calculator 1.7.0 1.7.1 1.7.2 1.7.3 giving me starkly different tCL values for the same timing when I "Import" my profile from Thaiphoon. To be exact, the "Import" feature says my kit can do 3733c14 at 1.42, vs even the "fast" feature of all those four versions show cl16
Starkly conflicting information such as one guy posting a GDM OFF CL15 build to copy (with missing voltage information), then another saying "GDM Off = Nuclear Physics, don't do that", yet can't even get it stable at CL16. Being linked to
own "guide" where the DEV himself admits on page 2, he has not tested a single REV-E kit himself ... (which is clear considering the outputs his program spits out as shown above)
At the end of the day, this isn't rocket science (as evidenced by thousands of people doing it). It very well may be a hardware limitation on my end (as
pointed out, I have a WorkStation board) -- but that same WS board is rated at 4400MHz, so, I don't know!
In short, I know people are helping (or at least trying to), I just think that if they kept in mind that I am a beginner when using their vast experience when providing advice, this would actually have resulted in some form of progress.
After 80 hours, the only thing that was "Stable" (outside of XMP) was 1 set of timings that didn't boot 85% of the time, and then told to try 40 other combinations at random doesn't help.
There has to be a logical process map that someone should be able to follow to get this done, I am not going for world records here, literally just trying to get something, anything, better than XMP that's stable, and have failed miserably at that.
Thanks for your perspective