Overclocking an FX-6350 with a 212Evo & 970A-UD3P?

[Pyroven]

My Build:

 

Gigabyte 970A-UD3P Motherboard, 8 pin ATX CPU power

Corsair CX600 PSU

Hyper212 Evo CPU Cooler

AMD FX-6350

 

Other:

16GB Crucial RAM

1TB HDD Seagate Barracuda

240GB Kingston SSDNow 300V

Fractal Design Define R4

 

 

I've been looking for a good guide on the internet about overclocking the FX-6350, but most are either for the 6300 or just generic guides that confuse me, who has never overclocked before.

 

 

What I think I know:

FX-6350 musn't go over 70*C

Increase core clock in BIOS 100MHz at a time, then stress test for 1 hour each time.

Silicon lottery means I may/may not get a crap OC.

 

I need to know:

What is that program that shows all the CPU info/running info that everyone seems to use?

Am I supposed to have some sort of drivers for my processor?

What are the risks?

How much will this affect the longevity of my processor?

Is my motherboard good enough for overclocking? It has an 8-pin 12V power thing for the CPU which I thought was supposed to be perfect for anything

How far can I push the temps until it gets dangerous? So if 70*C is bad, then should I stop at 60*C?

[OnekillWalter]

speedfan is nice, cpuz as well. speccy is more extensive. and i use Prime95 for stress test.

 

my fx-6300 is now at 4.4mhz and is stable at 53 celcius for an hour of torture test.

 

its mostly an trial/error procedure.

 

you must adjust the voltage of the cpu and the multiplier frequency. 

[Pyroven]

speedfan is nice, cpuz as well. speccy is more extensive. and i use Prime95 for stress test.

 

my fx-6300 is now at 4.4mhz and is stable at 53 celcius for an hour of torture test.

 

its mostly an trial/error procedure.

 

you must adjust the voltage of the cpu and the multiplier frequency. 

 

Right ok, that helps a bit.

[sgloux3470]

imo

 

Stress test for a max of 15 minutes.  Don't worry about temperatures, unless it throttles or shuts off it doesn't matter for a short stress test.  You only have to worry about thermal degradation if you're running those temperatures for long periods of time.  

 

Use real world testing (games, rendering etc) to determine overall stability and temperatures.  You may reach 75 degrees in prime95, but won't touch 60 in an actual game.  You're not going to run prime95 for the next year or two or whatever, but you'll probably play a lot of games.  Ensuring it is stable in real world scenarios means it will be stable and safe when you use it, while excessive stress testing will just reduce the lifespan of your components and only really tell you how well your room circulates air once your heatsinks/radiator are equilibrated with the temperature of your components.  

 

Only increase voltage if stress tests produce errors or you experience system crashes in real world usage.  You want the lowest voltage possible for a good overclock.  If you have the choice between +200 MHz w/ridiculous heat output and voltage and a lower overclock but massively better temperatures and lower voltage, choose the latter.  (Especially on an air cooler that will get dusty over time.  Here's what my hyper212 looked like after a little over a year:

 

You could probably start off aggressively with something like 4.0 GHz on stock voltage and then go slower from there.  

 

As for monitoring, HWMonitor and Core Temp are good.  

[Priller]

Just start raising the multiplier on a 6350 until it get unstable. My sample will do 4.7Ghz on stock voltage with 16GB of ram at 1866.

[WaitForMeImOnDialUp]

raising multiplier is the easiest way to overclock...but raising fsb instead increases lots of things mainly single core speed

 

i also agree when overclocking use hwinfo for monitoring and use intel burn test for stability.

 

your motherboard actually has a 8+2 power phase which is surprising for a 970 board. you should get good oc with that board but not sure with your cooler the most stable i could get with a 212 evo is 4.5ghz but thats with an 8350.

[Guest DollarPC]

imo

 

Stress test for a max of 15 minutes.  Don't worry about temperatures, unless it throttles or shuts off it doesn't matter for a short stress test.  You only have to worry about thermal degradation if you're running those temperatures for long periods of time.  

 

Use real world testing (games, rendering etc) to determine overall stability and temperatures.  You may reach 75 degrees in prime95, but won't touch 60 in an actual game.  You're not going to run prime95 for the next year or two or whatever, but you'll probably play a lot of games.  Ensuring it is stable in real world scenarios means it will be stable and safe when you use it, while excessive stress testing will just reduce the lifespan of your components and only really tell you how well your room circulates air once your heatsinks/radiator are equilibrated with the temperature of your components.  

 

Only increase voltage if stress tests produce errors or you experience system crashes in real world usage.  You want the lowest voltage possible for a good overclock.  If you have the choice between +200 MHz w/ridiculous heat output and voltage and a lower overclock but massively better temperatures and lower voltage, choose the latter.  (Especially on an air cooler that will get dusty over time.  Here's what my hyper212 looked like after a little over a year:

 

You could probably start off aggressively with something like 4.0 GHz on stock voltage and then go slower from there.  

 

As for monitoring, HWMonitor and Core Temp are good.  

Don't end up like him...Start low then go high go for turbo clock speed later start with stock

[Technology]

Hey I have same board and same processor and also using a hyper212evo.  I'm interested to bringing this thread alive.  I'm looking for a response before I discuss anything more.

[itachipirate]

My FX-6300 is at 4.0ghz easily on a 212evo and a motherboard with fantastic VRM cooling. If you want to play something like battlefield most people suggest a 4.3ghz overclock which isn't bad at all for these processors. My highest temp with all 6 cores at 4.0ghz through a torture test was 41 degrees celcius

 

Also, think of the FX-6300 and FX-6350 as the same thing. They're almost the same except the stock speed for the 6350 is faster

[Technology]

Yea, I played some of the Hardline beta and everything seem very smooth for me.  I can remember what Ghz I was when I played hardline though.

 

Right now I'm at 4.5Ghz and I have my loadline set to extreme.  I did have my CPU voltage set at the default setting but just now I have put the CPU voltage one notch below the default value.  Running prime95 and looking at my cpu voltage in HWMonitor it's staying around 1.476V.

 

I guess you could say that I'm more comfortable knowing my CPU will not go past 1.5V sense I don't believe 4.5Ghz really needs 1.5v.

 

When I had my CPU volts one notch above that default value and I ran prime95 and looked at my CPU volts in HWMonitor my CPU was easily hitting 1.5v and that was with my loadline set on extreme.

 

As i'm typing this now my CPU volts is showing 1.488v.

 

I think I'm going to leave it at 4.5ghz but I'm going to continue to lower the CPU volts unless or until prime95 starts failing.

 

I did have my loadline set on medium running 4.4ghz and I had my CPU volts a few notches above normal, when I 1st ran prime95 everything seemed to be stable and all around my cpu volts seem to stay low.  Then the other night I ran prime95 for giggles and I was getting failures.  So I quickly got upset and just went ahead and set my loadline to extreme and started running 4.5Ghz but started to lower my CPU volts sense on the extreme loadline the CPU volts seems to ramp up very easily.  

 

I really would enjoy to have any help of any kind, this is my 1st system I built for myself so I'm not very experienced with O/C. 

 

I have seen online that 1.5V on the CPU is not really a danger zone or anything, but I'm trying to feel comfortable with knowing my O/C is safe for 24/7.  Sometimes I do not log out of the game I usually play.   

 

*update* I just got a failure running Prime95 so I went back in bios and set the CPU voltage at the default value now HWMonitor is showing cpu volts at 1.488v instead of the previous 1.476v even though it was hitting 1.488v.  Now it's at 1.488V and I doubt it will drop any lower like at 1.476v.  All I really know to do now is just keep my loadline set on extreme and just keep uping the CPU volts if I see any prime95 failures.  If I start to get uncomfortable with the CPU volts I'll just lower back down to 4.4ghz and see what CPU volts I can have with that.  But yea now looking my CPU volts is dropping to 1.476v *lol* This is really time consuming for me if I get another failure on prime95 I'll up the CPU volts again. 

 

My question is 4.5Ghz what is the normal CPU volts for 4.5Ghz, as it is it looks like 1.476v - 1.5v with 1.5v looking to be the very stable voltage for me.  I'm looking for a 24/7 O/C and my temps are safe which they are at 45c right now so temps are not an issue or anything.  I guess you could say I maybe would feel better if I knew 1.5v for the cpu is ok and for 24/7 use. 

[miagisan]

i use for my daily driver 4.5ghz, 1.404 v, at load running occt it peaks at  55c after an hour. I have the exact same setup (212 eve and 970 ud3p). Keep it below 1.5v (according to general concensus of OCers) is fine for daily use. You should be fine at 1.5v if you keep the temps down below 62

[Technology]

Ah ok see that what I was wondering and all if 1.5V was OK if the temps are OK and stuff. So far i'm going to hold onto 4.5Ghz it's just that having the loadline set on extreme my volts usually don't go under 1.440V for the CPU at idle.  But I have my windows performance manager set to "performance" also if that makes a difference with what I seeing from the CPU volts and stuff

[Technology]

hey, it's been a while sense I've posted here.  I want to say that sense the last time i've posted here I've had my FX6350 at 4.92Ghz and upped the voltage to 1.600v but I was too scared to keep rasing the cpu volts and I could not get it stable with Prime95 and my CPU temps was getting too high like 70c and I noticed that I was having tempature spike or something when my CPU temp would just have a high spike and then maybe lower down to around 69c or so.  Right now my FX6350 is showing 1.524v CPU volts by looking at HWMonitor and I've had my FX-6350 clocked at 4.72GHZ for a long time now and prime95 runs stable.  I also did some overclocking with the FSB I believe 2003mhz was something I tried to achieve with my memory from the stock 1866MHZ.  I didn't really know what to do though when I reached a limit with the FSB, ...  I didn't know what to change with the FSB other than just backing the FSB multiplier back down whenever it was not stable. 

Raising the FSB should require what else to be changed?  That's what I'm looking to find out, is it ok for me to raise my memory above the stock 1.5v?  How do I know or how can I figure out what is keeping me from having a higher FSB overclock?

   As it is I backed my FSB back down and mainly just used the CPU multiplier to raise my CPU clock while also raising the cpu voltage.

When O/C the FSB,  should the voltage for the memory be raised any?  1.5v is my stock voltage for 1866mhz memory.

How about increasing other voltages like the north bridge or south bridge, ...  In the BIOS I leave mostly everything unchanged besides my CPU multiplier and my CPU core voltage.  I really try to find out what I can by looking on the internet, I would like to know if it's possible to have a stable memory overclock, is it OK to go past 1.5v for the memory?

 

My CPU temp stays around 60c-61c running Prime95 for like an hour and more with how my system is now.  I didn't mentioned that but I wanted too so I figured I would add that note for anyone that reads this post/thread/topic.

[Nena Trinity]

No idea if your mobo is a good OC board or not but if it is and you are sure then you can push it a bit more probably to 4.5 max?

[i_build_nanosuits]

I need to know:

-What is that program that shows all the CPU info/running info that everyone seems to use?

-Am I supposed to have some sort of drivers for my processor?

-What are the risks?

-How much will this affect the longevity of my processor?

-Is my motherboard good enough for overclocking? It has an 8-pin 12V power thing for the CPU which I thought was supposed to be perfect for anything

-How far can I push the temps until it gets dangerous? So if 70*C is bad, then should I stop at 60*C?

-HWiNFO64

-No

-Not much if you stay within 1.45v

-Not much if you stay within 1.45v

-Yes it is so far as you stay within 1.45v

-65c under full load

You're welcome, if you experience double post on boots after overclocking settings are applied on this board learn to deal with it or just update to the latest bios.

Good luck!