fx-8320 @4.5Ghz is 1.520V safe?

[cowpker4life]

so pretty much i want to know weather or not this is a safe voltage my cooling is fine with all cores pinnd it hits around 60-63c and it seems to be running some what stable running cpu-NB voltage at 1.392v and 2600mhz

with HT link running at 2400mhz. is there anyway to bring the voltage down and still have a stable overclock? im using a msi 970 gaming motherboard with a hyper 212 evo

[gtx1060=value]

NOTHING AT 1.5 VOLTS IS SAFE - THAT'S THE DEATH ZONE

[cowpker4life]

2 minutes ago, gtx1060=value said:

NOTHING AT 1.5 VOLTS IS SAFE - THAT'S THE DEATH ZONE

ive seen posts that 1.5-1.55v is okay for a fx-8xxx chip tho.

[SpaceGhostC2C]

It doesn't seem safe on that motherboard. It doesn't sound realistic on that cooler. 

Now, you say you somewhat pulled it off. If so, all you can do is try lower voltages and run stability tests. 

 

Coming to the idea itself, I have many doubts on how you are measuring things, but the 212 can't really keep up with 8 cores getting anything near 1.5v under full load, and the VRMs on the board may literally catch fire if stressed like this long enough. 

[cowpker4life]

using a core temp and burnintest. what do u suggest to do to try and get it to stabilize at lower volts?

[SpaceGhostC2C]

2 minutes ago, cowpker4life said:

ive seen posts that 1.5-1.55v is okay for a fx-8xxx chip tho.

Most BIOS will tell you a max voltage, which in my AM3+ boards is 1.545v.

Still, the problem isn't voltage, it's the current. At high enough voltages you push so much current to the CPU that either it or must likely the motherboard gives up if not prepared for it. 

[bcguru9384]

its only safe if your not directly to wall volts

wall volts fluctuate means clock will change also

it only takes a moment to lose cpu

lower volts be happy

[cowpker4life]

okay then. im going to keep looking into some beacause found a post on tomshardwere

 

The 8320 is clocked lower than the 8350. 

@OP, I would say around 1.45v should be your safe limit. You probably won't need that much. 
1.5v is where things start to get really detrimental to chip health, even with a suitable cooler, as far as I can remember something like 1.5v or 1.55 is the stated limit. 
 

[cowpker4life]

Just now, bcguru9384 said:

its only safe if your not directly to wall volts

wall volts fluctuate means clock will change also

it only takes a moment to lose cpu

lower volts be happy

so to not be directly from the wall volts what would it take? some type of converter?

[Ezio Auditore]

Noooooooooooooo!!!!!

And eben if your cpu is stable now it will die on booting as booting is nothing but votage spike. Clear CMOS before booting even rebooting

[cowpker4life]

so is there any way to get it stabilized then with lower volts

[bcguru9384]

3 minutes ago, cowpker4life said:

so to not be directly from the wall volts what would it take? some type of converter?

want a ups type powersupply

you can actually hook a car battery if it could transform 12v to 5v and 3v to your system then use a regulated charger on battery

very stable voltage then

but not easy project

but nearly no chance of volt spike occuring due to 120 shrunk to 12volt

regulators are great today but still run a ±15% volt range tolerance

[cowpker4life]

Just now, bcguru9384 said:

want a ups type powersupply

you can actually hook a car battery if it could transform 12v to 5v and 3v to your system then use a regulated charger on battery

very stable voltage then

but not easy project

but nearly no chance of volt spike occuring due to 120 shrunk to 12volt

regulators are great today but still run a ±15% volt range tolerance

ok ill have to pass on that. just dropped the clock back to 4.2 and back to auto voltage what should place it around 1.44 or something.

[bcguru9384]

that sounds like sound plan

happy computing

[cowpker4life]

1 minute ago, bcguru9384 said:

that sounds like sound plan

happy computing

thanks. just wish there was a way to get 4.5Ghz stable without 1.5v or more. but a lot of places im reading say that max voltage is 1.5-1.55v

[cowpker4life]

also how the hell is this cpu/mobo not burnning up at the voltages in the vid. its around 1.63v

 

[cowpker4life]

2 hours ago, wrathoftheturkey said:

Lol. On that mobo, chances are you're literally going to fry the VRMs before the chip burns out, but it'll be quite the race

MSI has a feature that is cpu smart protection. would this start to kick in and throttle the cpu if VRMs got to hot?

[Ryan Leech]

I've run 1.6v through my FX-6350 while doing a 5.4GHz run one time and it lived. I've been running my FX-6350 at 5.0GHz with 1.5v for 3.5 years now and it's still problem free. Although I have it under a water cooler and I'm using the GA-990FXA-UD5 which has a much better VRM setup than your MSI board.

 

I'd say your chip can handle the voltage just fine but your motherboard and cooling may not cope for 24/7 usage.

[cowpker4life]

2 minutes ago, Ryan Leech said:

I've run 1.6v through my FX-6350 while doing a 5.4GHz run. I've been running my FX-6350 at 5.0GHz with 1.5v for 3.5 years now and it's still problem free. Although I have it under a water cooler and I'm using the GA-990FXA-UD5 which has a much better VRM setup than your MSI board.

 

I'd say your chip can handle the voltage just fine but your motherboard and cooling may not cope for 24/7 usage.

the cpu smart proctection should help to keep the VRM from burning up tho right? or nah

[Ryan Leech]

1 minute ago, cowpker4life said:

the cpu smart proctection should help to keep the VRM from burning up tho right? or nah

CPU smart protection is just going to throttle the cpu which would diminish the point of overclocking in the first place. I looked into your motherboard and the VRM doesn't look all that bad, although with those HOT Nikos mosfets you'll want another fan blowing on your VRM heatsink.

[cowpker4life]

you think it would be safe to leave it at that without having a fan on the VRMs? i dont have the cpu pinned most the time its under 50% usage but that is because games only pin 1-2 cores. is there any software to manage the temps of the socket and VRMs? and if so what are the safe temps.

[cowpker4life]

does increasing cpu-nb or ht link help stabilize the cpu?

[FiveNine]

So far it just seems like you're trying to get people to agree or to ok what you're doing. Probably for the best to dial it down imo, i would not be gambling. That does not seem like an impressive overclock for that voltage to be worth it either...

[SpaceGhostC2C]

On 3/4/2017 at 6:03 AM, cowpker4life said:

the cpu smart proctection should help to keep the VRM from burning up tho right? or nah

You can't trust it with your life, though:

Spoiler

 

(The 9590 is essentially your same CPU clocked at 4.7Ghz and vocre set above 1.5v)

11 hours ago, cowpker4life said:

does increasing cpu-nb or ht link help stabilize the cpu?

Minimally. They may help you get more stable results at high RAM speeds (your CPU runs cooler if memory is running slow, as the integrated memory controller stays cooler). If your RAM is running at the intended speed, I'll leave it at that (I think you increase the CPU-NB voltage: you can probably dial it back closer to stock). Another possibility is to run your RAM at a standard speed (1600, for example) if its XMP is high. Not sure whether the OC it's worth using slower RAM, though.

[PovilJ]

I have fx-8320, I broke 1 pin and it didn't work through it inside mobo and it actually worked. Disabled turbo and now I am stable running 4.2ghz at 1.44V. So as a broken CPU i am happy.