TPU's Core i7 Overclocking and Feedback

[Wile E]

Ok, based on what I asked that is the max output amplitude of their VRMs, and the output voltage of the VRMs is x1.05 the VCore. The tech may have been an idiot... that is totally possible. But each phase of a power system is rated for a certain ammount of amps. The new EVGA classified is open with their specs and their total output current is 400amps at 40amps per phase at 10 phases... if this has 12 phases at 11 amps per phase (rounding) then it makes sense.

If I write them another e-mail, what should I ask instead? I'm very interested in finding this information.

Just point out what I did. That 130A can be taken up by a purely stock chip if it has a higher VID, and that 130A doesn't seem right.

I wonder if he meant 310A?

 

[PaulieG]

Just point out what I did. That 130A can be taken up by a purely stock chip if it has a higher VID, and that 130A doesn't seem right.

I wonder if he meant 310A?

I would bet that he did mean 310A.

 

[Binge]

I am resending the e-mail with that question. I am also asking for their white sheet. An e-mail has also been sent to DFI, ASUS, and MSI.

Just make sure you guys think about this before buying your x58. What are their components actually rated for? Are they forcing you to drastically push the motherboard as well as the CPU if you want to overclock?

 

[PaulieG]

I am resending the e-mail with that question. I am also asking for their white sheet. An e-mail has also been sent to DFI, ASUS, and MSI.

Just make sure you guys think about this before buying your x58. What are their components actually rated for? Are they forcing you to drastically push the motherboard as well as the CPU if you want to overclock?

I think this would be a good question to post over at Rebels. Polygon seems to be the Biostar guru, similar to Rgone or Praz at DFI. I have the Biostar on it's way, and it should be here tomorrow, and I'll certainly push it from day one.

You may also want to post the question on DFI Club.

 

[Binge]

Have done. The DFI club guys told me to contact their tech support? lol

 

[kid41212003]

An e-mail has also been sent to DFI, ASUS, and MSI.

No GIGABYTE UD5 or XTREME?

 

[freakshow]

what is the max volts to use on the 920's?

 

[Fitseries3]

thats how you need to tune a phase/cascade if any of you build one. just figure 15% over OCed wattage.

 

[Fitseries3]

what is the max volts to use on the 920's?

gah!

everyone stop asking that please!

i have run up to 1.65v and its been fine AS LONG AS YOU HAVE GOOD COOLING.

 

[SystemViper]

me too

 

[PaulieG]

gah!

everyone stop asking that please!

i have run up to 1.65v and its been fine AS LONG AS YOU HAVE GOOD COOLING.

Fit, it's a legitimate question. I'm sure you've heard of electron migration. You CAN kill a chip due to this, and it has nothing to due with heat. At some point, the voltage itself will degrade to EM. This is fact.

 

[freakshow]

gah!

everyone stop asking that please!

i have run up to 1.65v and its been fine AS LONG AS YOU HAVE GOOD COOLING.

lol i wasnt asking u lmao i was asking whoever is willin to answer


anyways sorry

 

[Fitseries3]

not a big deal really.

no one is going to run 1.65v 24/7 so yeah....

 

[Binge]

Newcomers should read some of the 60+ pages of good info we have on here along with screenshots. If they want intel specs, they should ask Intel how much juice they approve in spec.

 

[Wile E]

Fit, it's a legitimate question. I'm sure you've heard of electron migration. You CAN kill a chip due to this, and it has nothing to due with heat. At some point, the voltage itself will degrade to EM. This is fact.

I have to agree. I would bet money that 1.65V is not safe for extended periods of time, even with sub-zero cooling. No amount of cooling will stop EM. EM is why Yorkfields degrade when using over 1.4V, even with good cooling.

 

[SystemViper]

Still
-THIS IS MY OPINION-

I think these chips can handle some more voltage then even the 1.65v without damage if it's some benching and not 24/7,

I always look at the the super clockers are doing on LN2 and they are hitting it with much more voltage. So that if short spurts for benchers I don't think it will degrade the chip. But also Cold changed the speed and other stuff at the molecular level so going LN2 actually creates a situation that can't be replacated any other way, cold water, regular watercooliong or even dice.



that being said, for most pople "don't try this at home" and stay within Intel spec's

 

[PaulieG]

Well I've gotta say... I would appreciate it if newcomers would read some of the 60+ pages of good info we have on here along with screenshots. If they want intel specs, they should ask Intel how much juice they approve in spec.

The real answer we need will most likely take some time, and unfortunately some dead chips. We need to know how much above spec vcore is still relatively safe with good cooling, while minimizing the effects of electron migration.

 

[SystemViper]

I have to agree. I would bet money that 1.65V is not safe for extended periods of time, even with sub-zero cooling. No amount of cooling will stop EM. EM is why Yorkfields degrade when using over 1.4V, even with good cooling.

I have a different opinion about running 1.4v on the Yorkies, I have some that have been running for a long time with no degradation, i'm not saying it won't happen, but 1.4v isn't even extreem

 

[Binge]

The real answer we need will most likely take some time, and unfortunately some dead chips. We need to know how much above spec vcore is still relatively safe with good cooling, while minimizing the effects of electron migration.

My point is this: For the longevity of your processor you must stay within intel spec because intel spec has the testing done that proves those are the safe operating voltages to protect against heat damage and EM. Relatively safe is a term my friends and I like to use when we light off illegal fireworks. It doesn't mean that one of them won't blow our hand off no matter how "relatively safe" we've been previously.

 

[SystemViper]

how do you test for "against heat damage and EM" put it under an electron microscope~ lol, jsut kiding but wonder how you would really tell, prob not something a layman could do

 

[Fitseries3]

its relatively safe to sit in an office chair but one out of 10000 has to have a problem that anally kills who ocupied it with metal shards...

http://i.gizmodo.com/5157354/boy-killed-anally-when-office-chair-explodes#c

 

[kid41212003]

I remembered somewhere, that if you're running at spec 24/7, an Pentium III could last you at least 10 years. I'm wondering if it's the same thing with the newer CPUs .

 

[SystemViper]

 

[kid41212003]

I'm sure his own gas triggered that explosion.

 

[PaulieG]

how do you test for "against heat damage and EM" put it under an electron microscope~ lol, jsut kiding but wonder how you would really tell, prob not something a layman could do

What we need is data over time to see how many people kill their chips, and at what vcore they were using. This is what's occured with every generation of chips. It's how we came to accept 1.5v as a safe max for C2D chips, and 1.45v for yorkies for 24/7 use. There are numerous threads with reports that helped to support this to a reasonable extent.