Athlon 3700+ - 1.48 vcore... Is this normal??

Hendrix89

Hi, ive just noticed on CPU.Z recently that my CPU voltage is at 1.488v! I checked in the BIOS and sure enough, Vcore = 1.48. From doing some research ive noticed that the average and safe vcore voltage for an athlon CPU is 1.35v!? I did do a small Overclock on my CPU recently (2.2 to 2.4Ghz) but i did not tamper with or increase the vcore/voltage... Its wierd cause i have not had any problems at all with my system so far and its running as stable as can be... Are these readings normal?? I need to know soon so i can do something about it.

Hendrix89

Look i really need help ppl! I only need an answer....

jozi

My winchester 3200 is at 1.36V why dont you just put the voltage down in bios?

As they say in dutch "gedult is een schoone zaak"

Hendrix89

Yeah but what if thats (1.48) the default/stock voltage for my athlon 3700+? I mean i have never touched the voltages...EVER so how could they change by themselves. Thats why i need to be absolutely sure that its really is not right for my CPU before i mess with voltage...

Energizer

Ye, my amd64 3700 is showing a vcore of 1.41.

I never overclocked it. Is this safe as i am having problems with the machine?

Hendrix89

Energizer wrote:

Ye, my amd64 3700 is showing a vcore of 1.41.

I never overclocked it. Is this safe as i am having problems with the machine?

Well its not as high as mine but if your having problems then it thats a bit odd cause my system is running fine and seems alright appart from my vcore voltage, which is much higher than yours... This is actually a good oppertunity to compare. By any chance have you recently flashed your BIOS to latest version? I did that recently (Asus premium mobo) and im starting to think that the new version raised the voltage. Any one else out there with an Athlon 3700+ or similar, would be helpful if you could state your default voltage.

Cuddlesworth

It took me 6 months to kill an Athlon at 1.8volts, so dont worry.

Hendrix89

krazy_8s wrote:

It took me 6 months to kill an Athlon at 1.8volts, so dont worry.

Alright, but do you think that with my unusually high stock CPU voltage i have a lot more OCing room, without needing to raise the vcore from where it is now at all? cause the max im probly gonna go (when i get my new cooler) is 2.8Ghz. I may not even go that high... 2.6 for awhile anyway. I mean there has to be a benefit.

Cuddlesworth

No, as far as I remember Amd never gave a set stock voltage for any Venice or San deago chips, because some needed more volts at stock than others.But it is unasually high, and the fault then lies with the mobo. If your worried put it down to 1.43, thats if you intend on keeping your chip for the next few years.

Hendrix89

krazy_8s wrote:

No, as far as I remember Amd never gave a set stock voltage for any Venice or San deago chips, because some needed more volts at stock than others.But it is unasually high, and the fault then lies with the mobo. If your worried put it down to 1.43, thats if you intend on keeping your chip for the next few years.

Ah crud... im worried now! I really do not feel comfortable changing the voltage at all so i dont know what to do!!? Ill probly keep this CPU for another year and a half, 2 yrs at the most so do you think itll be alright for that time without any problems?? I will be OCing more during that time though.... Also do you think maybe its just the BIOS version i have? maybe when i flash to a new one the voltage problem will be fixed...?

L31mr0d

whats the big problem? So let me get this straight you are having no problems with the cpu, no heat problems, in fact your computer is running perfectly stable and is set at stock voltage settings yet somehow you've convinced yourself this is bad? Have you looked up what the max voltage for your CPU even is? I think its around 1.7v as I have the same one.

Listen the only reason voltage causes your CPU to fail is that increasing it generates more heat (resistance, electrons, size, blah di blah), eventually your CPU will burn itself out if the voltage is too high for too long. I've got my 3700+ (san diago) @ 2.8Ghz @ 1.45v and the temps max at 55'c. Check your temps, a stock 3700+ @ 2.2Ghz should max at 35 at best to 45 depending on your stepping and luck

Synopsis: Voltage good, temperature bad. If your CPU idles at 20'c at 1.9v then leave it there.

Cuddlesworth

On the safe side 1.55volts for air
1.6-1.7 for water
1.8 volts upwards for phase.

Doesnt matter about heat, you can damage a cpu by higher volts, but at your setting your looking at about 3 years at least before you see problems.

Hendrix89

L31mr0d wrote:

whats the big problem? So let me get this straight you are having no problems with the cpu, no heat problems, in fact your computer is running perfectly stable and is set at stock voltage settings yet somehow you've convinced yourself this is bad? Have you looked up what the max voltage for your CPU even is? I think its around 1.7v as I have the same one.

Listen the only reason voltage causes your CPU to fail is that increasing it generates more heat (resistance, electrons, size, blah di blah), eventually your CPU will burn itself out if the voltage is too high for too long. I've got my 3700+ (san diago) @ 2.8Ghz @ 1.45v and the temps max at 55'c. Check your temps, a stock 3700+ @ 2.2Ghz should max at 35 at best to 45 depending on your stepping and luck

Synopsis: Voltage good, temperature bad. If your CPU idles at 20'c at 1.9v then leave it there.

Okay just to let you know my 3700+ is at 2.4Ghz. My idle temps are bout 33/34C and load is in the low to mid 40's. I understand what you mean but the thing is your stock voltage was obviously much lower than mine even is at stock(at the moment) and even now with yours at 2.8 your voltage still does not need to be as high as mine needs to be to simply handle 2.4Ghz... I barely understood what i said there myself! So what do you think? Also next time i want to OC (say to 2.6 from 2.4) will i have to raise my already high voltage up more??

L31mr0d

krazy_8s wrote:

Doesnt matter about heat, you can damage a cpu by higher volts, but at your setting your looking at about 3 years at least before you see problems.

eh, what do you think raising the voltage does that damages the cpu if not heat. I did 5 years of electronic engineering and for semiconductors with transistors and circuits on the micron scale, heat and ESD are your only concerns (next to cosmic rays... wtf!!!). Why do you think they say you've "fried" that IC. Pushing electrons through silicon gives off heat as a biproduct, if you push too many electrons through (i.e. raising voltage) they either begin to bunch up at areas of greater resistance or break down the gaps between connections by means of the excessive heat generated.

My mobo defaulted my CPU to 1.46v initially, I OC'd to 2.8Ghz and it ran prime for 12 hours stable, I lowered the voltage to see if I could maintain that clock at a lower voltage and I could, it dropped my max temp from 59'c max to 55' which i'm happy with. If you are worried that it is too high then just lower it, run some stress tests (OCCT or prime95) if it doesn't give a BSOD or restart (set your system to restart if your CPU hits 60'c in your BIOS) then its stable. An overclocked CPU will last you 2 years usually. Its the people that want to keep their computer for a decade that shouldn't overclock.

Hendrix89

L31mr0d wrote:

eh, what do you think raising the voltage does that damages the cpu if not heat. I did 5 years of electronic engineering and for semiconductors with transistors and circuits on the micron scale, heat and ESD are your only concerns (next to cosmic rays... wtf!!!). Why do you think they say you've "fried" that IC. Pushing electrons through silicon gives off heat as a biproduct, if you push too many electrons through (i.e. raising voltage) they either begin to bunch up at areas of greater resistance or break down the gaps between connections by means of the excessive heat generated.

My mobo defaulted my CPU to 1.46v initially, I OC'd to 2.8Ghz and it ran prime for 12 hours stable, I lowered the voltage to see if I could maintain that clock at a lower voltage and I could, it dropped my max temp from 59'c max to 55' which i'm happy with. If you are worried that it is too high then just lower it, run some stress tests (OCCT or prime95) if it doesn't give a BSOD or restart (set your system to restart if your CPU hits 60'c in your BIOS) then its stable. An overclocked CPU will last you 2 years usually. Its the people that want to keep their computer for a decade that shouldn't overclock.

Okay i guess if you did it safely then i might try lowering the vcore a bit. What do you suggest i could lower it to (safely)?

SouperComputer

Okay i guess if you did it safely then i might try lowering the vcore a bit. What do you suggest i could lower it to (safely)?

Lower it as much as you want, you arent gona do damage. Use stress prime to test its stability after dropping the voltage.

Knock .2V off and see how it goes.

L31mr0d

AMD 64 Processor Power and Thermal Datasheet

If you go to page 39 of this datasheet (or do a search for 3700 in it) you'll see that the max (ie worst case) temp range for this CPU (chip code: ADA3700AEP5AR) is between 49'c to 71'c at a voltage of 1.35 to 1.4v @ 2.2Ghz.

On page 33 it shows that at 2.4Ghz the CPU should be at 1.5v. Drop it to 1.45 and test it there, then up your clock from there if you want to go further. I wouldn't drop it below 1.4 as you are already running an OC on the CPU.

Cuddlesworth

to L31mr0d, heat is a concern as any overclocker knows, but over time electron migration occurs. The time in which this can occur is increased by overvolting the cpu.

Hendrix89

L31mr0d wrote:

AMD 64 Processor Power and Thermal Datasheet

If you go to page 39 of this datasheet (or do a search for 3700 in it) you'll see that the max (ie worst case) temp range for this CPU (chip code: ADA3700AEP5AR) is between 49'c to 71'c at a voltage of 1.35 to 1.4v @ 2.2Ghz.

On page 33 it shows that at 2.4Ghz the CPU should be at 1.5v. Drop it to 1.45 and test it there, then up your clock from there if you want to go further. I wouldn't drop it below 1.4 as you are already running an OC on the CPU.

Okay then on second thought i might leave the vcore as it is then for now. There is something wierd going on in my BIOS though... when i went to check out what i could set my vcore to there were no options!? I selected vcore in the BIOS and the only only option was ''Ignored". Unless thats not where i set the voltage...? I might have to raise the voltage in the future when i OC again so i need to get this sorted out...

L31mr0d

krazy_8s wrote:

to L31mr0d, heat is a concern as any overclocker knows, but over time electron migration occurs. The time in which this can occur is increased by overvolting the cpu.

So your saying that if you increase the clock, which isn't increasing the voltage that EM wouldn't happen. Increasing the clock or the voltage causes EM to happen due to the heat they produce. EM is the result of damage on a semiconductor circuit, it is not the cause. The best way to explain it is to use a meandering river as an example, if the flow of water is too great it will eventually erode the sides of river. This erosion is the resistance of the electrons against the silicon, what wears the silicon is the heat caused by this friction. Explain, without factoring heat, how you believe increasing voltage causes Electron Migration?

awhir

krazy_8s wrote:

It took me 6 months to kill an Athlon at 1.8volts, so dont worry.

your crazy!!!!!!!!!!!!:D or would it be krazy

Cuddlesworth

L31mr0d wrote:

So your saying that if you increase the clock, which isn't increasing the voltage that EM wouldn't happen. Increasing the clock or the voltage causes EM to happen due to the heat they produce. EM is the result of damage on a semiconductor circuit, it is not the cause. The best way to explain it is to use a meandering river as an example, if the flow of water is too great it will eventually erode the sides of river. This erosion is the resistance of the electrons against the silicon, what wears the silicon is the heat caused by this friction. Explain, without factoring heat, how you believe increasing voltage causes Electron Migration?

I cant.

Redisle

Ive got my 3700 in a epox 9npa ultra board and the v core is always between 1.48 and 1.51. That shows on a little epox tool called ghost monitor. Reading this thread has got me a bit worried because I have never oc/d the cpu or messed with the voltages or anything so is my vcore normal? My cpu is at 2.21 ghz.

Also is it normal for the vcore to constantly change?

L31mr0d

Check what the voltage in the BIOS is at a few boots to see if it really is changing. I use systool personally and it seems to fluctuate its reading of the voltage between 1.44 - 1.47, but in the BIOS it is always the same at 1.45. So it might be just an issue with the epox software making an approximation

If you've read this thread you'll know that if you are not having heat problems with the voltage settings as they are then there shouldn't be a problem. But, to be safe, i'd lower the voltage, to maybe 1.45 and run some stress tests there, the lower the better, and the reduction in heat will give you room to try some overclocking if you want.

zilog_jones

It's probably just the slight inaccuracies of the power supply. The Vcore voltage on any PC I've looked at has always been slightly off - usually higher than what it's supposed to be. Anywhere within 10% of the standard voltage is probably OK.