Vcore and vdimm with A64

**don_vercetti** · 11-10-2005, 11:02 AM

it's hardly AMD's fault. I'm sure they don't recommend using more than the JEDEC max to the RAM, so that's what, 2.85v?

It does seem that a lot of the deaths have been people running high volts through the RAM (3.4v+) while keeping their fx-57s at stock voltage. At least we know what's causing it now, and people should stop having these troubles.

**Kameleon** · 11-10-2005, 11:06 AM

Interesting. I'll have to spread the word about this. Could this also be a cause of instability on all A64s when pushing the vDIMM up to ~4v (what I tend to do with my BH-5/booster), do you need to stick the vCore up above 1.6v or so?

Something I'll have to look out for if I decide to pull this BH-5 and booster out for fun with the Opteron...it only needs 1.35v to max out on my board, so I could've easily fried it with 4v through the RAM.

**mouzel** · 11-10-2005, 12:22 PM

If you use C&Q you can run 1.0 V Vcore and 3.0 Vdimm for instance. That's a difference of 2.0 Volt and works fine. Maybe the voltage differential between Vcore and Vdimm may not exceed 2.5 V or something like that ?

**racinjimy** · 11-10-2005, 12:25 PM

Originally Posted by mouzel

If you use C&Q you can run 1.0 V Vcore and 3.0 Vdimm for instance. That's a difference of 2.0 Volt and works fine. Maybe the voltage differential between Vcore and Vdimm may not exceed 2.5 V or something like that ?

right so don't use C&Q if overclocking..........that is in every AMD overclocking sticky on every forum

**xgman** · 11-10-2005, 12:44 PM

So has anyone dtermined if this thread is a result of this issue?

http://www.xtremesystems.org/forums/...7&page=1&pp=25

**TheMeatFrog** · 11-10-2005, 02:19 PM

I run my bh-5 at 3.4v or 3.5v most the time so my vcore should be atleast 1.45 correct? Now is this 1.45v windows vcore or bios?

**RyderOCZ** · 11-10-2005, 02:28 PM

I would say bios or "calculated"... i.e. 1.325 + 110% = 1.4575...but that will show probably 1.41-1.43 in bios? Even if you went 1.350 + 110...that should put you at 1.45 in bios I think.

**TheMeatFrog** · 11-10-2005, 02:52 PM

Wow thanks for the fast reply Ryder!

**T_M** · 11-10-2005, 03:43 PM

Interesting that AMD suggested raising vcore

**Tony** · 11-11-2005, 05:51 AM

Originally Posted by T_M

Interesting that AMD suggested raising vcore

They didn't, they would never say that officially.

The only way to strengthen the core is with a vcore increase, thats all they will actually say.

**Jochenp** · 11-11-2005, 04:47 AM

Lol, nice one T_M

they suggest you void your warranty! I'm not gonna do that

**Kameleon** · 11-11-2005, 07:25 AM

So can we get an answer on this: What's a safe vCore to run if you're running 3.5-4V vDIMM?

**RyderOCZ** · 11-11-2005, 10:09 AM

Originally Posted by Kameleon

So can we get an answer on this: What's a safe vCore to run if you're running 3.5-4V vDIMM?

There is no Spec for this...nobody gave Tony or anyone else any definitive numbers. The only thing AMD officially says is:

Originally Posted by Tony

The only way to strengthen the core is with a vcore increase, thats all they will actually say

So just don't run Stock Vcore with High Vdimm....MY OPINION and best guess would be at least 0.05 over stock...maybe 0.1V

So if stock Vcore is 1.40...run 1.45-1.50V.

**Monge** · 11-11-2005, 10:24 AM

RyderOCZ: Very clear, thanks

Would you consider all past 3v or 3.2v as high Vdimm?

**RyderOCZ** · 11-11-2005, 10:33 AM

Originally Posted by Monge

RyderOCZ: Very clear, thanks

Would you consider all past 3v or 3.2v as high Vdimm?

IMO yes 3.0V+ is high Vdimm.....whether 3.0V and using stock Vcore is enough to harm your CPU....who knows.......but yes I consider 3.0V+ high Vdimm

**Takumi** · 11-11-2005, 10:52 AM

Tony: do you know what the reverse breakdown voltage range is for these diodes. Knowing this could help narrow down the range of safe voltage differences.

**Micutzu** · 11-11-2005, 11:31 AM

First of all, the CPU doesn't have to do directly with the VDD, but with Vtt/Vref, wich should vary from 1.25 to 2V when you adjust the RAM from 2.5 to 4V. Maximum JEDEC Vdd voltage is 2.85V, and lowest voltage for A64 CPU's (CnQ activated) is 1.1V, so we know for sure that a difference of 0.325V between Vcore and Vtt is safe. This means that for a 1.35V Vcore voltage, we can have a safe Vdd of at least 3.35V, and for 1.5V Vcore -> 3.65V Vdd.
Probably the real danger is when those diodes go near the opening (conduction) voltage, wich should be ~0.6V for Si-based junction; if this is true, 1.35V Vcore and above 3.9V Vdd should kill the CPU instantly ... anyone wanna try ?

**Zeus** · 11-11-2005, 12:20 PM

Interesting stuff.

A while back i bought a second hand venice 3000+ 0512 that could do 2,7GHz at stock voltage.
When the former owner first got it it was able to do 290+ Mhz with TCCD, later on he switched over to redline and high voltage.

After a few months of "abuse" with high voltage i bought the CPU and tried to get it stable at 9x300 with my TCCD.

There was no way to get it stable over 260 any more.
2.7GHz with stock voltage was still possible though.

I think the memcontroller got worn out due to high voltage.

Fortunately, the seller was kind enough to let me return the CPU.

**Tony** · 11-12-2005, 06:27 AM

Originally Posted by Micutzu

First of all, the CPU doesn't have to do directly with the VDD, but with Vtt/Vref, wich should vary from 1.25 to 2V when you adjust the RAM from 2.5 to 4V. Maximum JEDEC Vdd voltage is 2.85V, and lowest voltage for A64 CPU's (CnQ activated) is 1.1V, so we know for sure that a difference of 0.325V between Vcore and Vtt is safe. This means that for a 1.35V Vcore voltage, we can have a safe Vdd of at least 3.35V, and for 1.5V Vcore -> 3.65V Vdd.
Probably the real danger is when those diodes go near the opening (conduction) voltage, wich should be ~0.6V for Si-based junction; if this is true, 1.35V Vcore and above 3.9V Vdd should kill the CPU instantly ... anyone wanna try ?

you are touching on what I have tried to tell every one. its not the value of VDD, its the voltage difference between VDD and vcore on the cpu you all need to watch and keep as small as possible.

**ozzimark** · 11-13-2005, 10:40 PM

Originally Posted by Micutzu

First of all, the CPU doesn't have to do directly with the VDD, but with Vtt/Vref, wich should vary from 1.25 to 2V when you adjust the RAM from 2.5 to 4V. Maximum JEDEC Vdd voltage is 2.85V, and lowest voltage for A64 CPU's (CnQ activated) is 1.1V, so we know for sure that a difference of 0.325V between Vcore and Vtt is safe. This means that for a 1.35V Vcore voltage, we can have a safe Vdd of at least 3.35V, and for 1.5V Vcore -> 3.65V Vdd.
Probably the real danger is when those diodes go near the opening (conduction) voltage, wich should be ~0.6V for Si-based junction; if this is true, 1.35V Vcore and above 3.9V Vdd should kill the CPU instantly ... anyone wanna try ?

that's good logic.. however, that would imply that 3.45v is fine with 1.4v vcore (3.45/2 = 1.725 - 1.4v = .325)

remember that power draw increases with the square of the voltage

**high5** · 11-14-2005, 01:50 AM

Originally Posted by ozzimark

remember that power draw increases with the square of the voltage

in theory yes. real world - not quite.

**ozzimark** · 11-14-2005, 05:44 AM

well yes, esp since silicon has an interesting habit of decreasing resistance, thus increasing current as temps go up. instead of going into the complexities of it all, i think making a rough guide by leaving it simple would work better

however, with that in mind, it would make sense that ram like micron, that runs hot because it draws more power, would also be a higher strain on those diodes at a given voltage, no? and that a single sided stick won't be nearly as bad as a double sided stick, and that 4 dimms will be worse than 2. there's just a hell of a lot of variables in this.

**wittekakker** · 11-15-2005, 11:45 AM

Originally Posted by Micutzu

Probably the real danger is when those diodes go near the opening (conduction) voltage, wich should be ~0.6V for Si-based junction; if this is true, 1.35V Vcore and above 3.9V Vdd should kill the CPU instantly ... anyone wanna try ?

Some more details on how this works would be nice, I like to understand it 100%.
It makes much sense

**Vassili** · 11-21-2005, 07:53 AM

Originally Posted by Micutzu

First of all, the CPU doesn't have to do directly with the VDD, but with Vtt/Vref, wich should vary from 1.25 to 2V when you adjust the RAM from 2.5 to 4V. Maximum JEDEC Vdd voltage is 2.85V, and lowest voltage for A64 CPU's (CnQ activated) is 1.1V, so we know for sure that a difference of 0.325V between Vcore and Vtt is safe. This means that for a 1.35V Vcore voltage, we can have a safe Vdd of at least 3.35V, and for 1.5V Vcore -> 3.65V Vdd.
Probably the real danger is when those diodes go near the opening (conduction) voltage, wich should be ~0.6V for Si-based junction; if this is true, 1.35V Vcore and above 3.9V Vdd should kill the CPU instantly ... anyone wanna try ?

Someone went very close:

XtremeSystems Forums > Xtreme > Xtreme Bandwidth > @ My Kingston @ 280 2-2-2-5 8M stable @ 3,68v

Strange, Oh and don't get me wrong here, I do believe you Tony