First look at Piledriver, how does it compare.

[Hits9Nine]

Well, I know that Vdroop enabled prevents voltage from sagging upon load. You can reach the same clocks without it by just adding vcore. LLC does not have anything to do with reaching the 'true value' of a chip AFAIK.

For example say 4.5Ghz 1.45v load is stable and you have .03 vdroop on load. You can:

1. Set bios to 1.48v with NO LLC and it will droop to your load voltage.
2. Set bios to 1.45v with LLC and it will stay at 1.45v.

6 of one, half dozen of the other there at the same clocks... doesnt matter what clockspeed though.

The point is LLC doesnt unleash anything but more stable voltages upon load. I don't imagine this to be different on BD as it was on all Intel chips I have overclocked. This is the first time I heard of it increasing potential of a CPU and I'm having trouble believing it I must admit (because that isnt true truth be told).

But not having LLC and setting the cpu volt to 1.6v to achieve 5.0+ghz oc on the BD for 10 min is dumb and very unstable, or can set LLC and have 1.46v and have 5.0+ ghz STABLE ...you save money, have less heat and get better OC's too. I am sorry keeping the FX8120 stable at 5.0ghz gets pretty awesome results. AM3+ MoBo without LLC good luck keeping BD stable at 4.6ghz

Term stable= not crashing under load, running Prime without errors for 7+hours, not 5 min.

 

[EarthDog]

1.45v with LLC is the same as 1.45v without LLC when you are talking load voltage. No money is saved, no heat is lost, no overclock is better(higher). If you do not use LLC and use vcore to compensate, your idle voltage will be higher than it needs to be. Thats all.

 

[hokiealumnus]

Hits9Nine, I'm sorry but I think you have a fundamental misunderstanding of what LLC does, precisely. As we're trying to explain to you, LLC reduces Vdroop, or eliminates it altogether. Perhaps an example would better explain it.

If you set (random guess) 1.52Vcore in BIOS with LLC off, it will probably achieve the same 5.0GHz stable overclock, because Vdroop will cause the Vcore to droop to 1.46V under load and be just as stable as your overclock.

By enabling LLC, you can set Vcore at 1.46V and it will remain at 1.46V when loaded, because the LLC setting you're chosen makes loaded Vcore the same as idle Vcore, thereby eliminating Vdroop.

Depending on the board and particular setting, LLC can actually overshoot voltage, simply give less Vdroop or keep the voltage constant. It's a feature added to give us extra control of how the voltage is applied to our CPUs. The lack of that feature does not, by itself, keep people from "know[ing] the true speed of a Bulldozer." As long as loaded Vcore is the same, your chip should do just as well in a board without LLC. It will just require a higher idle Vcore than you are using currently.

 

[Hits9Nine]

1.45v with LLC is the same as 1.45v without LLC when you are talking load voltage. No money is saved, no heat is lost, no overclock is better(higher). If you do not use LLC and use vcore to compensate, your idle voltage will be higher than it needs to be. Thats all.

So with an FX8120 for me to get 5.0+Ghz without LLC I would have to set Vcore to 1.6v+ and its not 100% stable, does that seem smart to you?
Yeah lets just jack up the voltages so we can hit 5.0ghz for 15-20 min and fry our CPUs, that sounds like a great idea!

Again BD chip without LLC on the MoBbo can't achieve great oc's without jacking up the voltage. Having LLC helps achieve better overclocks, maintaining stability and not pushing the voltages beyond stupidity making the BD preform better and stable.

 

[Hits9Nine]

Hits9Nine, I'm sorry but I think you have a fundamental misunderstanding of what LLC does, precisely. As we're trying to explain to you, LLC reduces Vdroop, or eliminates it altogether. Perhaps an example would better explain it.

If you set (random guess) 1.52Vcore in BIOS with LLC off, it will probably achieve the same 5.0GHz stable overclock, because Vdroop will cause the Vcore to droop to 1.46V under load and be just as stable as your overclock.

By enabling LLC, you can set Vcore at 1.46V and it will remain at 1.46V when loaded, because the LLC setting you're chosen makes loaded Vcore the same as idle Vcore, thereby eliminating Vdroop.

Depending on the board and particular setting, LLC can actually overshoot voltage, simply give less Vdroop or keep the voltage constant. It's a feature added to give us extra control of how the voltage is applied to our CPUs. The lack of that feature does not, by itself, keep people from "know[ing] the true speed of a Bulldozer." As long as loaded Vcore is the same, your chip should do just as well in a board without LLC. It will just require a higher idle Vcore than you are using currently.

Oh I understand alright, with out vdroop LLC ON I can oc my BD better and more stable, with it off I have to jack up the voltage and risk frying my cpu and have a less stable system

 

[EarthDog]

DELETED.. Quoting Hokie.. well said.

Hits9Nine, I'm sorry but I think you have a fundamental misunderstanding of what LLC does, precisely. As we're trying to explain to you, LLC reduces Vdroop, or eliminates it altogether. Perhaps an example would better explain it.

If you set (random guess) 1.52Vcore in BIOS with LLC off, it will probably achieve the same 5.0GHz stable overclock, because Vdroop will cause the Vcore to droop to 1.46V under load and be just as stable as your overclock.

By enabling LLC, you can set Vcore at 1.46V and it will remain at 1.46V when loaded, because the LLC setting you're chosen makes loaded Vcore the same as idle Vcore, thereby eliminating Vdroop.

Depending on the board and particular setting, LLC can actually overshoot voltage, simply give less Vdroop or keep the voltage constant. It's a feature added to give us extra control of how the voltage is applied to our CPUs. The lack of that feature does not, by itself, keep people from "know[ing] the true speed of a Bulldozer." As long as loaded Vcore is the same, your chip should do just as well in a board without LLC. It will just require a higher idle Vcore than you are using currently.

i will add that if you (Hits) have significant vdroop and are not being dramatic to try and prove your point, you have a motherboard problem and need to RMA it or get a better motherboard. Can you list your ACTUAL voltages (set in bios w/out LLC and with + windows idle voltage, then windows load voltage with LLC and without).

I think you may have a board problem which is shaping/causing this misconception of LLC and what it REALLY does for you.

 

[Hits9Nine]

DELETED.. Quoting Hokie.. well said.



i will add that if you (Hits) have significant vdroop and are not being dramatic to try and prove your point, you have a motherboard problem and need to RMA it or get a better motherboard. Can you list your ACTUAL voltages (set in bios w/out LLC and with + windows idle voltage, then windows load voltage with LLC and without).

I think you may have a board problem which is shaping/causing this misconception of LLC and what it REALLY does for you.

AM3+ mobos have vdroop issues with BD chips, I have used Gigabyte 990fx ud7, MSI 990fx-g80 and Asus Crosshair V, once Gitabyte ud7 and Asus Corsshair have LLC enabled they work great, without LLC there are large Vdroop under load causing instability and poor oc's, google Bulldozer Vdroop you will see

 

[EarthDog]

I did. They enable LLC, like they are supposed to, problem resolved. For most situations though its not a big enough problem to warrant a statement that LLC improves peak overclocking because, well, it doesnt. It can go either way. It saves a higher idle voltage is all. If you need 1.45v to be stable, it doesnt matter if LLC is on or off, you will be stable at 1.45v. And when you arent running really high 24/7 clocks at 5Ghz like you say you are, you have that option to go either way.


Oh well, enough about this... back to the PILEDRIVER!

 

[Hits9Nine]

DELETED.. Quoting Hokie.. well said.



i will add that if you (Hits) have significant vdroop and are not being dramatic to try and prove your point, you have a motherboard problem and need to RMA it or get a better motherboard. Can you list your ACTUAL voltages (set in bios w/out LLC and with + windows idle voltage, then windows load voltage with LLC and without).

I think you may have a board problem which is shaping/causing this misconception of LLC and what it REALLY does for you.

Read this, if you want.
http://www.overclockersclub.com/reviews/gigabyte_990fxaud7/6.htm

I know its a different site, it goes along with what I was saying BD and 990FX Vdroop and LLC

 

[EarthDog]

Sounds like a problem with the board to me (and my first google result)... the voltages shouldnt be jumping around like that without LLC...

Thanks for the link.. Generally however, LLC doesnt help like that. There are always exceptions to the rule though.

 

[Hits9Nine]

I did. They enable LLC, like they are supposed to, problem resolved. For most situations though its not a big enough problem to warrant a statement that LLC improves peak overclocking because, well, it doesnt. It can go either way. It saves a higher idle voltage is all. If you need 1.45v to be stable, it doesnt matter if LLC is on or off, you will be stable at 1.45v. And when you arent running really high 24/7 clocks at 5Ghz like you say you are, you have that option to go either way.


Oh well, enough about this... back to the PILEDRIVER!

Yes back to Piledriver, I can't wait... I am itching to build another PC

 

[BeepBeep2]

Hits9Nine, while I understand what you are saying to some extent you are making a hyperbole out of reality.

Lets say I have an FX-8150 CPU that needs 1.45v under load to pass Prime95 for...4 hours. Stable.
On a revision 1.0 UD7, I would need 1.575v to reach 1.45v at load. 1.575v is unsafe, therefore this is a horrible option. You had that correct.
On a revision 1.1 UD7, I would need 1.47v to reach 1.45v at load with LLC Enabled and set to "Ultra High". This is wonderful, much safer, less stupid and great.

However each time I need 1.45v to reach the same frequency, and I don't save any power, and do not overclock any higher. The power is not cleaner, nor does the voltage that comes at load with LLC have any kind of magic fairy dust.

I think you are a bit confused here.

CPU-Z validations ("5 GHz + unstable at 1.6v") which is taken at idle (<1% CPU load) will come at 1.6v regardless of LLC being on at not. The only difference is that when a load is applied without LLC, then the voltage will droop when it will stay at 1.6v under load with LLC adjusted correctly. (In which case your CPU will probably die rather quickly, or at least degrade over time due to electromigration related to higher voltage than the internal circuits can handle.)

By internal circuits, I am really dumbing it down, I am talking about the microscopic circuits that make up a CPU that are also in this case 32nm in diameter in various layers of the "die" in the CPU


Furthermore, until you prove otherwise, you are telling a lie about your CPU being stable at 5 GHz+ for 7 hours unless using chilled water or some other kind of exotic cooling.

That's like me claiming I have a 4.7 GHz stable 965 BE on air...

Make sense?

 

[Hits9Nine]

Hits9Nine, while I understand what you are saying to some extent you are making a hyperbole out of reality.

Lets say I have an FX-8150 CPU that needs 1.45v under load to pass Prime95 for...4 hours. Stable.
On a revision 1.0 UD7, I would need 1.575v to reach 1.45v at load. 1.575v is unsafe, therefore this is a horrible option. You had that correct.
On a revision 1.1 UD7, I would need 1.47v to reach 1.45v at load with LLC Enabled and set to "Ultra High". This is wonderful, much safer, less stupid and great.

However each time I need 1.45v to reach the same frequency, and I don't save any power, and do not overclock any higher. The power is not cleaner, nor does the voltage that comes at load with LLC have any kind of magic fairy dust.

I think you are a bit confused here.

CPU-Z validations ("5 GHz + unstable at 1.6v") which is taken at idle (<1% CPU load) will come at 1.6v regardless of LLC being on at not. The only difference is that when a load is applied without LLC, then the voltage will droop when it will stay at 1.6v under load with LLC adjusted correctly. (In which case your CPU will probably die rather quickly, or at least degrade over time due to electromigration related to higher voltage than the internal circuits can handle.)

By internal circuits, I am really dumbing it down, I am talking about the microscopic circuits that make up a CPU that are also in this case 32nm in diameter in various layers of the "die" in the CPU


Furthermore, until you prove otherwise, you are telling a lie about your CPU being stable at 5 GHz+ for 7 hours unless using chilled water or some other kind of exotic cooling.

That's like me claiming I have a 4.7 GHz stable 965 BE on air...

Make sense?

This is all I have for you at the moment,
http://valid.canardpc.com/show_oc.php?id=2533160 , and its done with an H100 last night, just for you!

My Swiftech H20-220 Ultima HD is not setup (which did the initial Prime) as its not fitting into my case the way I wanted it and I don't want to use the green cooling as it looks ugly... waiting for the white stuff to arrive. So no Exotic cooling here, but thanks for calling me a liar. Anyway I don't care if you don't believe me, I'm not trying to impress you! Also by the time I get my white cooling for the Swiftech H20-220 Ultima HD, and figure out how to make it look good in my Raven I will have Piledriver and will then Prime the Piledriver at 5.0ghz hopefully.

 

[BeepBeep2]

lol, this has nothing to do with you impressing me, this has to do with you backing up your claims.

If I saw a CPU run P95 stable at 5 GHz and 1.46v, I'd probably pay several hundred dollars for it.
Either you discovered more than a gem or don't know what you're doing...

Making a CPU-Z validation at 1.49v just under 5 GHz is not nearly the same.

 

[Hits9Nine]

lol, this has nothing to do with you impressing me, this has to do with you backing up your claims.

If I saw a CPU run P95 stable at 5 GHz and 1.46v, I'd probably pay several hundred dollars for it.
Either you discovered more than a gem or don't know what you're doing...

Making a CPU-Z validation at 1.49v just under 5 GHz is not nearly the same.

I really don't care what you think, or what your getting at. I don't need to prove anything to you! I know what I can do and what I have! Sorry I don't Prime again for 7 more hours. I have a job and family with kids that I have to tend to, and when I have time I spend it very efficiently with my PCs. So setting up my other water cooling system and running Prime95 for 7 more hours just to get you off my back is a big waste of my time since I don't have much time in the first place.
I don't believe you know what your talking about, you will have to prove it... now how about you prove it to me! To me your just some dude on the internet who thinks he knows what he is talking about...see how I can play this game too! No matter what you say to me I wont believe you either, so I think this conversation is done now. BYE

 

[hokiealumnus]

But....he's not the one making the claim of 5GHz stable at 1.46V, nor is he the one making over-reaching, exaggerated claims about LLC. You shouldn't really be upset when people ask for some semblance of proof of such claims, at least IMHO.

 

[Hits9Nine]

But....he's not the one making the claim of 5GHz stable at 1.46V, nor is he the one making over-reaching, exaggerated claims about LLC. You shouldn't really be upset when people ask for some semblance of proof of such claims, at least IMHO.

LMFAO...over-reaching claims! Did you read the article?

http://www.overclockersclub.com/revi...90fxaud7/6.htm

Hmmmm! same Issue I was going through with no LLC!

I see maybe I will stop posting on this site and just stay there from now on.

 

[hokiealumnus]

Yes, I read that article. In fact, I reviewed the 990FXA-UD7. They had sent me a Rev. 1.0 board, which was horrible at voltage control and did not have LLC. The problem isn't totally that it didn't have LLC (though LLC does help smooth things out on that board), it's that the voltage wasn't constant. Even without LLC, the voltage should remain constant when under load, but at a lower level due to Vdroop. Obviously that's not the case as you can see from their graph, so on Rev. 1.1 boards, Gigabyte introduced LLC that smoothed out the voltage. It was a bandaid to patch a problem, but never fixed the original problem. Voltage should not fluctuate like that with LLC on or off.

So yes, you're making over-reaching comments. You're also making unsubstatiated claims about overclocks. If your metric for posting here is that people just take your word for everything, then perhaps your choice is for the best.

 

[Hits9Nine]

Yes, I read that article. In fact, I reviewed the 990FXA-UD7. They had sent me a Rev. 1.0 board, which was horrible at voltage control and did not have LLC. The problem isn't totally that it didn't have LLC (though LLC does help smooth things out on that board), it's that the voltage wasn't constant. Even without LLC, the voltage should remain constant when under load, but at a lower level due to Vdroop. Obviously that's not the case as you can see from their graph, so on Rev. 1.1 boards, Gigabyte introduced LLC that smoothed out the voltage. It was a bandaid to patch a problem, but never fixed the original problem. Voltage should not fluctuate like that with LLC on or off.

So yes, you're making over-reaching comments. You're also making unsubstatiated claims about overclocks. If your metric for posting here is that people just take your word for everything, then perhaps your choice is for the best.

did you turn LLC off on the rev1.1 and get the same overclock? I still get large vdroop with rev 1.1 with llc off

 

[hokiealumnus]

No, it wouldn't reach the same overclock with LLC off on that board. That's why I said it was a band-aid to patch the problem, not a fix. They never fixed the larger problem in the first place - that their power section was incapable of keeping voltage constant. Instead they introduced the controller that had LLC and left the actual power delivery the same. The LLC chip controls voltage much better than the board does by default, allowing people to use it to smooth out the voltage output.

LLC should be there to control Vdroop (or eliminate it), it should not be there to fix a problem of voltage output fluctuating dramatically.