- Thread Starter
- #61
Confused. I set the vcore voltage to 1.425000 as you know, was that the MAX allowed and not how much it is actually getting or something? Otherwise how am I gonna add more juice if 1.425000 is already the max I should go?
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
Looking for help with first CPU overclock (FX-4170).
- Thread starter Obsidus
- Start date
Confused. I set the vcore voltage to 1.425000 as you know, was that the MAX allowed and not how much it is actually getting or something? Otherwise how am I gonna add more juice if 1.425000 is already the max I should go?
- Thread Starter
- #63
Ok! I keep trying to run Prime95, and it lasts varying amounts of time, though never longer than a minute. Keeps giving me:
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
On various different cores, one core at a time and so far not multiple failures at once. Is it possible to give it more voltage than it needs, and therefore make it unstable? As in, maybe it doesn't even need 1.42500 to run at 4.6? I suppose not, seeing as how that message is telltale of not enough juice eh.
- Thread Starter
- #64
Hmm just read something that said that this problem is usually with the RAM or something to that effect. They said to either bump the NB voltage or the ram voltage. I'll try to run the other tests on P95, and if they finish then maybe it is the ram since blend is supposedly hard on the ram.
- Joined
- Dec 27, 2008
The ram wouldn't keep you from not being able to increase the vcore beyond 1.425 and that is what you need to be able to do. That is the cause of your instability. Maybe someone else who has your same board can help with this issue. I suggest posting in the AMD motherboard section with this issue. Give it a subject line like: "Help! My Sabertooth bios won't let me increase the vcore past 1.425".
- Thread Starter
- #67
Oh I think we have a misunderstanding here. I CAN increase it past 1.425000, it won't stop me, but I haven't done it because I was under the impression that it wasn't safe past 1.42500. Or maybe you did understand me and theres still a problem.
- Joined
- Dec 27, 2008
Why did you get the impression that it wasn't safe to increase the vcore beyond 1.425? Does the bios give you a warning or do the values turn red in color or something? If so, this is just a hedge and overclockers ignore those things. Again, temps are the thing to pay attention to. vcore values per se are not a concern until you significantly exceed 1.5 volts. Even 1.55 vcore under load is considered safe by most people.
- Thread Starter
- #70
I think I got that in my head from all the crap I've been reading all over google, trying to learn and psych myself up enough to finally take the dive and OC my cpu. Like I said, I'm paranoid about wrecking my computer somehow, so little things like that stick in my mind, even if they're wrong,
- Joined
- Dec 27, 2008
That tells you the need was more vcore. When you add more and it helps then that was what it needed. Keep an eye on temps.
- Thread Starter
- #75
Ok made it to test number 6 on blend, with all temps staying at 49c, but then it failed on worker 4 and gave me a different reason this time:
FATAL ERROR: Final result was 3E9BA002, expected: 33BB2689.
Hardware failure detected, consult stress.txt file.
More juice, or is this something different?
FATAL ERROR: Final result was 3E9BA002, expected: 33BB2689.
Hardware failure detected, consult stress.txt file.
More juice, or is this something different?
- Thread Starter
- #77
Last update for the day.
Have managed to run Prime95 blend for around 45 min with temps never going above 52c and usually hovering around 48-50c , at 4.6 and 1.465ish voltage. Everything seemed fine, but around 46 min in I noticed that the tests hadn't moved on for a while; turns out it didn't fail like before, but instead my whole computer locked up. I don't think I'm going to worry at all about trying for 4.7+ and instead am gonna focus on trying to get 4.6 completely stable. Dunno if the computer completely freezing up is a sign I need more voltage, or if I've pushed it too far at this point.
Will be back tomorrow.
Have managed to run Prime95 blend for around 45 min with temps never going above 52c and usually hovering around 48-50c , at 4.6 and 1.465ish voltage. Everything seemed fine, but around 46 min in I noticed that the tests hadn't moved on for a while; turns out it didn't fail like before, but instead my whole computer locked up. I don't think I'm going to worry at all about trying for 4.7+ and instead am gonna focus on trying to get 4.6 completely stable. Dunno if the computer completely freezing up is a sign I need more voltage, or if I've pushed it too far at this point.
Will be back tomorrow.
- Joined
- Dec 27, 2008
I think you're okay for one small bump in vcore. One thing to realize is that the 4170 is already overclocked some from the factory so it may not overclock any higher than the 4100. It's mostly a marketing strategy aimed at non-overclockers who want a little faster chip. I think you are wise to work at getting it stable at 4.6. Good night.
Last edited:
- Thread Starter
- #79
So I woke up this morning and tweaked the voltage to 1.475 and ran blend again. It lasted for about 45 min before worker 3 had that damned 0.5, expected 0.4 error again. My temperatures are not getting any higher than 51 on the cpu and 53 on the cores (and it usually drops down to between 48-51) so I want to say it isn't heat causing instability. I'm starting to wonder if this can be solved by adding more voltage, or if I'm SOL and need to lower my clock at this point. Or maybe I need to adjust the ram or overclock the NB or something?
- Joined
- Dec 27, 2008
One thing you probably should look at is the ram timing. Unfortunately, CPU-z sometimes gives wrong info in the SPD tab about the mfg.'s recommendations. There's a bug in it, or some versions of it, that causes it not to read the mfg.'s info correctly from the ram module. So, go to your ram manfuacturer's website and check the timings and voltage recommendations of the ram. Then go into bios and put that menu option for DRAM timing on Manual so you can check what the current timings are set to. Compare them with what the mfg recmmends. In bios, you will see a ton of different timings but the only ones you need to be concerned with are the ones that show in CPU-z. The ones that don't show up in CPU-z are not very important and you can leave them on Auto.
Check and see what the DRAM voltage is set to as well.
If you find the terminology in CPU-z for some of the timings isn't the same as it is in bios and get confused, take a digital camera pic of that section of the bios and I'll try to help you sort it out. You can often tell just by the magnitude of the numbers what lines up with what when the terminology isn't the same.
Check and see what the DRAM voltage is set to as well.
If you find the terminology in CPU-z for some of the timings isn't the same as it is in bios and get confused, take a digital camera pic of that section of the bios and I'll try to help you sort it out. You can often tell just by the magnitude of the numbers what lines up with what when the terminology isn't the same.
