Asus Sabertooth 990fx R2.0 and AMD FX 8350 OC help needed

[Nenad]

Yes you'll want to set it to the 2133 profile but you see where it sayd 11-13-13-13-48 shouldbe more like 11-13-13-36-48. Set the trfc to 300 again. It looks pretty obvious to me tha the board is NOT reading the profile on this memory properly.

And right now we need to try this with the lower CPU_NB the board supports 2400(OC) that's what that means it has to be OCd to run that speed.
The F12 works but only if your flash has been formatted in FAT 32 not NTFS


EDIT Is it runing that ram 1t 2145 with the 11-11-11 timings?? If so that's fine.

Okay going to set it to 11-13-13-36-48 and trfc is at 300 already. So you want me to run a prime test with that memory settings and to leave it at current 2145 MHz? What is next step after it?
And cheers for the F12 tip, i formated the USB now to FAT32

 

[Johan45]

Well IF you want to run at 2400 we have to up the NB speed so that will be our next hurdle. After seeing the readings in the bios I can see why these kingston kits are so messed up. Theyre not reading the profiles correctly so the bios sets it up all wrong.

I just went back and checked your CPU-z and the ram setting at that speed is 11-13-13-30-48 1T

 

[Nenad]

Okay I got the memory settings to 11-13-13-36-48. So I'll run a test to see if all is okay here and if it is then you suggest me bumping up to 2452Mhz memory and the NB one bump up? And what should I put memory settings at? Since the one doesn't seem to be right again for 2400MHz. It says 11-13-13-7-54.

 

[Johan45]

We're not going to do both of them at once . We'll start with the NB see if we can make that work. As for timings refer to CPU-z spd tab it seems to be reading it correctly although you'll see less timings. You may have to get a more complete set of timings from kingston for your specific set of ram if we still have issues.

 

[Nenad]

Even if it is stable at this speed, I am kinda satisfied. There is good new Johan, tomorrow my case should be here. I got a email to expect the package tomorrow.

As for CPU speed i'd like when it is in the new case for us to bump it a bit higher towards 4.7GHz, max 4.8GHz. But yeah, lets go step by step and do the NB first. Prime is running now. To make sure these timings are not making any problems. I'll get back to you as soon as it is done x)

 

[Johan45]

OK . I'll have another hour , then I'll be out of communication for a while have to make a road trip to pick up my kid from University.

 

[Nenad]

No problem mate! I'll go and drink a bit since it is Friday night(already 7PM here). x) I'll let you know when it passes. Around 10mins more of the test left. x)

 

[trents]

My suspicion is Kingston has catered to the Intel crowd in their SPD info and so AMD chipsets and bioses do not interpret it correctly. This may be why there have been so many compatibility issues with Kingston on AMD systems in recent years. I wonder if end users could contact Kingston and get timings that are more appropriate for AMD. Of note is that AMD has begun marketing its own line of RAM specifically geared to AMD chipsets.

 

[Nenad]

My suspicion is Kingston has catered to the Intel crowd in their SPD info and so AMD chipsets and bioses do not interpret it correctly. This may be why there have been so many compatibility issues with Kingston on AMD systems in recent years. I wonder if end users could contact Kingston and get timings that are more appropriate for AMD. Of note is that AMD has begun marketing its own line of RAM specifically geared to AMD chipsets.

Only if I was smart enough and actually went on and asked you guys before buying my RAM what would be best a long side for my AMD system...
I guess people learn while they are alive. x)

 

[Johan45]

My suspicion is Kingston has catered to the Intel crowd in their SPD info and so AMD chipsets and bioses do not interpret it correctly. This may be why there have been so many compatibility issues with Kingston on AMD systems in recent years. I wonder if end users could contact Kingston and get timings that are more appropriate for AMD. Of note is that AMD has begun marketing its own line of RAM specifically geared to AMD chipsets.

I think that's wwhy I've been having more luck with oldr bios versions in these situations Trents. They just seem to read the ram more appropriately. If the PC had set up to that spd there's no way it would run.

 

[Nenad]

Johan, test passed in Prime after 25mins with these settings. Ready for next instructions. x)

 

[Johan45]

OK let's try that NB again at 25xx but before you do I want to bump the CPU_NB voltage to 1.28 see if that pulls it through.

 

[Nenad]

Everything was looking good, temperatures didn't raise and all was similar to the 23xx NB but then after 12min's Worker #2 stopped.

 

[Johan45]

Ah yes but we're making progress. I was hoping from the time span you were going to have good news and it's not all bad. At this point I would give the CPU just a bump in voltage Just 1 and see if that impoves things at all.
When we tried that NB speed earlier it crashed fairly quickly, didn't it.
If the CPU bump doesn't improve things we may need a bit more to the CPU_NB then. We're getting a bit higher there than usual but you do have 16 Gig of uncooperative ram there so it may be needed.

 

[Nenad]

Yeah, it instantly crashed after first few minutes. I'll do a bump on Vcore just a bit. Or should i maybe just put LLC a bump so it gives more voltage then bumping it down? x)

 

[Johan45]

No , I would just adjust the voltage. I'm taking off soon. I think you have a pretty good idea of what you're doing at this point. Don't be afraid to try things if this doesn't do anything add some CPUNB voltage but only a bit at a time now. If you try to get that ram up to 2400 make sure to save your profile before hand. But you do have one saved so you have a starting point if things go off track. Also at some point you'll want to try prime for a couple hours to get a good solid clock you can always fall back to.
Have fun tonight Nenad

 

[Nenad]

Yeah, will try what you said mate and i'll probably retire for tonight going out a bit. Thanks for all the tips, i'll post what i got from results for you.
Make sure you drive safe! Cya mate!

 

[Johan45]

Yep no probs , I'll be around tomorrow morning again for sure

 

[Nenad]

I gave my Vcore 1 and then 2nd bump, was at 1.425V and both times test failed early on in first 3mins, which i don't get since it was 12mins with 1.4V. I'll call it for tonight as going out, will pick it up when i get up x)

 

[Nenad]

Hey Johan! New case arrived and all is assembled and I am quite pleased with diffrence it made! I bumped my settings down to last stable to compare results and they are quite a eye candy! CPU temp went down from 64 degrees MAX to 56 degrees MAX and Package stayed same at around 50-51.
I am gonna post you the snap shots and the pictures i took with my phone so you can see how it looks now. x)

And here is the cooler on backplate of MOBO. I took your advice.

Gonna run Prime for 2h+ as you suggested to test stability and then we can go into conquering that NB and maybe up the CPU even more now that we have almost 10 degrees more space then when it was on my table. x)