-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
AMD FX Overclocking
- Thread starter LtMatt
- Start date
Great news Gamble!
To be honest unless your benching I found going above 4.5Ghz was wasted as it only produced more heat and with that instability chance for the 8320 of mine.
I can push it to 5Ghz but it is not stable, I think one core is the culprit and not worth the excessive voltage. 4.5Ghz plays most games just fine so I have mine around this frequency too.
Thanks all! I am only sitting at 4.6 due to the way I did the multiplier but will just stay here if you feel going higher doesn't add any extra real life performance.
I am using a Phanteks PH-TC14PE CPU Cooler which is an absolute beast!!! I get between 21 and 23 degrees idle with a max of 43 degrees on load. This translates to 8 degrees on the socket on idle and 29 degrees on load or thereabouts.
In addition to that CPU cooler, I have 2x120mm fans on the front of my case pulling in, 1x140mm exhaust at the rear, 2x140mm exhaust at the top, and a 200mm fan on the side pulling air in as well as a 120mm fan at the bottom pulling air in.
The above configuration is the best I have found so far for all my components as it results in a 50degree max temp on load for my two 770's.
I will check the max I get too later on, never thought to check that to be honest!
I got to disagree on the E chips Davidree; my 8320E hit 5 Ghz with just chaning the mulitplier and going to 1.45v....with ave 1.4 volts; once I started playing with northbridge and fsb I got 5.055 Ghz at 1.46 volts......
I could go higher but I think one of the modules is crappying now; most 8350s don't do 5Ghz easily or a lot higher volts; 9000 series can but not usually that low of volts.
I think the E chips are another stepping; or they've improved the process that much. I could back her back down to something like 4.8 and most likely be 1.35v or less but I kinda like having 5 ghz machine
Dude as I said I disagree and show me where I said fsb was HT?? I didn't meantion HT as I was at work and didn't remember the numbers off the top of my head
Fsb up top in your first post, and no Ht Reference is not the Ht link either. and how can you disagree if we are agreeing on the same thing you turnip.
Associate
- Joined
- 29 Sep 2011
- Posts
- 132
I had an 8350 that lasted at 4.8 for about 6 months. Was upset when the mobo died and took the cpu with it, now I have a replacement board without a cpu 
Wifey brought me a 4690k and to be honest I'm not overly happy with it, its a fraction better in gaming but I miss having the extra 4 cores.
AMD deserve a bit more credit than they get.
Wifey brought me a 4690k and to be honest I'm not overly happy with it, its a fraction better in gaming but I miss having the extra 4 cores.
AMD deserve a bit more credit than they get.
I had an 8350 that lasted at 4.8 for about 6 months. Was upset when the mobo died and took the cpu with it, now I have a replacement board without a cpu
Wifey brought me a 4690k and to be honest I'm not overly happy with it, its a fraction better in gaming but I miss having the extra 4 cores.
AMD deserve a bit more credit than they get.
What was the reason for the mobo dying? The OC or something else?
Associate
- Joined
- 29 Sep 2011
- Posts
- 132
I think it was just a faulty board from the start, Ive blamed it on everything but truth be told i dont know, it came rushed out of the factory straight to my door due to high demand, the thing didn't even have a serial number in the bios but the replacement did and it was trouble from the start.
I had put it back to stock to benchmark a game, as soon as I clicked open. It died instantly and never started again.
Thing is that's 2 MSI products in a year that have failed on me and I was a big fanboy of there products but now think twice before wanting to buy anything from them.
Oh dear not good. Were you able to do anything with warranties and have you replaced the MSI components?
So, hit problems almost straight away with mine .
First of all, my rev1 990fx sabertooth appears to be overvolting, quite a bit. With turbocore on, it sits at 1.42 in bios, which from what I can tell is the turbo core boost voltage. Under heavy load (realbench), that peaks at 1.44/5. Temps are still fine (55C) @4.2, but leaves little room for anymore. Turbo core off. Voltage in bios defaults to 1.4 and peaks at 1.42 in realbench.
It didn't overvolt my 955 x4, is it because it's a rev1 and needed a bios update?
Anyway, luckily the 990fx can apply a negative offset, and applying negative of 0.018 bring it down to 1.38 in bios and 1.4 under real bench. So I am going try and bring it down to 1.38 under load as that's suppose to be default and there go from there.
Is there a better way to stop it doing so? or is just negative offset the best way while still letting it clock down (fixed voltage won't let it).
.First of all, my rev1 990fx sabertooth appears to be overvolting, quite a bit. With turbocore on, it sits at 1.42 in bios, which from what I can tell is the turbo core boost voltage. Under heavy load (realbench), that peaks at 1.44/5. Temps are still fine (55C) @4.2, but leaves little room for anymore. Turbo core off. Voltage in bios defaults to 1.4 and peaks at 1.42 in realbench.
It didn't overvolt my 955 x4, is it because it's a rev1 and needed a bios update?
Anyway, luckily the 990fx can apply a negative offset, and applying negative of 0.018 bring it down to 1.38 in bios and 1.4 under real bench. So I am going try and bring it down to 1.38 under load as that's suppose to be default and there go from there.
Is there a better way to stop it doing so? or is just negative offset the best way while still letting it clock down (fixed voltage won't let it).
Last edited:
Soldato
Have you updated to the latest BIOS? Site has 1604 for 8320's.
I have 1604.
It's the only bios for rev1 boards that will even take an fx. I updated to it before I changed over to the fx.
Soldato
OK.
Did you take off the energy saving features and look at the voltage?
Did you take off the energy saving features and look at the voltage?
Soldato
So, hit problems almost straight away with mine
First of all, my rev1 990fx sabertooth appears to be overvolting, quite a bit. With turbocore on, it sits at 1.42 in bios, which from what I can tell is the turbo core boost voltage. Under heavy load (realbench), that peaks at 1.44/5. Temps are still fine (55C) @4.2, but leaves little room for anymore. Turbo core off. Voltage in bios defaults to 1.4 and peaks at 1.42 in realbench.
It didn't overvolt my 955 x4, is it because it's a rev1 and needed a bios update?
Anyway, luckily the 990fx can apply a negative offset, and applying negative of 0.018 bring it down to 1.38 in bios and 1.4 under real bench. So I am going try and bring it down to 1.38 under load as that's suppose to be default and there go from there.
Is there a better way to stop it doing so? or is just negative offset the best way while still letting it clock down (fixed voltage won't let it).
Have a look for an option called LLC (Load Line Calibration) in your BIOS - That might be what is causing the overvolting. Let us know what level LLC is set to.
Not relevant to his situation.
The correct advice as llc could be too high, also disable turbo boost as it increases the voltage a little, and a negative offset isn't a bad thing as long as your idle voltage isn't so low that the cpu becomes unstable.
Seems like it goes over by 0.02. With Turbo off, it's 1.4V in bios, 1.42 under load. Offset by -0.02 so it's 1.38 in bios and it's 1.4 under load. Not sure why it thinks 1.4V is default, when it's suppose to 1.38 though. It's the same even If I use a fixed 1.38V instead of offset.
Anyway LLC. LLC is set to Auto. It has 5 settings.
Auto
Regular (0%)
Medium (25%)
High (50%)
Ultra High (75%)
Extreme (100%)
K, I'll give high llc a go at the weekend (won't have time to mess around till then
)As for the other questions.
Alpenföhn K2 Mount Doom cooler
Asus Realbench
Package was max 55C, I forgot to write down what the socket was. I know it was not near max temp. I'll make sure to note both down next time so I don't forget.
K, I'll give high llc a go at the weekend (won't have time to mess around till then
As for the other questions.
Alpenföhn K2 Mount Doom cooler
Asus Realbench
Package was max 55C, I forgot to write down what the socket was. I know it was not near max temp. I'll make sure to note both down next time so I don't forget.
I also have the board and revision but don't have your problem, running an 8370. It def has its "quirks" though (probably why Rev 2 came out!). I seem to fail anything that goes over FSB 210. tried jumping it to see if its just a bit of a black spot and all sorts but to no avail.
I'm going to reseat my cooler soon I think as although its an older cooler there shouldnt be a 10C temp difference between yours and mine especially when I'm running at 1.35/1.37 and you are much cooler! you don't have a silly (10 fans etc) case at all do you?
Low under normal circumstances but not really under stress test. just on the edge. I'm going to reset my cooler I think as armadillo is using 0.05v more and getting 10C less!
my 4.7ghz @ 1.368 under load run.
Last edited: