Gigabyte Z97 Boards Teazer

[wazza300]

Re: Gigabyte Z97 Boards Teazer

that's usually down to loadline calibration boosting vcore slightly to counter vdroop under load

just adjust the loadline calibration level and or the cpu vcore offset (dvid) to suit and lower the load cpu voltage

[James7679]

Re: Gigabyte Z97 Boards Teazer

FWIW, an Intel tech told me that the voltage range 4790k, goes up to 1.5.
Sent from my Lumia 1520 using Tapatalk

[MysticalOS]

Re: Gigabyte Z97 Boards Teazer

I see the voltage do that too. I set mine to 1.258 and it actually does 1.272

I still have it at 4.6. Running at 70C under load on noctua D14. I upgraded to a XSPC Raystorm AX360 Extreme Universal CPU Water Cooling Kit w/ D5 Variant Pump and it still runs at 70C under load. I'm baffled as to why the raystorm isn't better. I was hoping to shoot for 4.8-5.0 with best water cooler i could get and turns out to break even with best air cooling out there :\ Locked at 4.6 it is I guess. :\

Taking it apart twice now I don't think I can fix the problem, i think that all around the block is poorly designed. it's so large because it's supposed to be a multi cpu block. The problem is, the socket bracket is taller than the cpu, so I think there is too big of a pocket between the block and the cpu itself because of the bracket. the noctua has a much smaller block that is only the size of cpu, so it sits on and presses against cpu more firmly. as good as water cooler is, it's no good if it's not tight enough. for a 300 dollar product you'd think they'd design a block that is better and isn't cooling the bracket assembly instead of the cpu. Unfortunately i can't record a product for being crappy.

[scracy]

Re: Gigabyte Z97 Boards Teazer

that's usually down to loadline calibration boosting vcore slightly to counter vdroop under load

just adjust the loadline calibration level and or the cpu vcore offset (dvid) to suit and lower the load cpu voltage

To the best of my knowledge you cant adjust LLC on Gigabyte Z87 boards,at least not for the Vcore...you do have one for Vrin though....

[Lui]

Re: Gigabyte Z97 Boards Teazer

I have a new (delidded) 4790K, which can make 4.6Ghz 1.150V, 5Ghz IBT stable 1.387V..... I can't upload pictures

[boondocks]

Re: Gigabyte Z97 Boards Teazer

I see the voltage do that too. I set mine to 1.258 and it actually does 1.272

I still have it at 4.6. Running at 70C under load on noctua D14. I upgraded to a XSPC Raystorm AX360 Extreme Universal CPU Water Cooling Kit w/ D5 Variant Pump and it still runs at 70C under load. I'm baffled as to why the raystorm isn't better. I was hoping to shoot for 4.8-5.0 with best water cooler i could get and turns out to break even with best air cooling out there :\ Locked at 4.6 it is I guess. :\

Taking it apart twice now I don't think I can fix the problem, i think that all around the block is poorly designed. it's so large because it's supposed to be a multi cpu block. The problem is, the socket bracket is taller than the cpu, so I think there is too big of a pocket between the block and the cpu itself because of the bracket. the noctua has a much smaller block that is only the size of cpu, so it sits on and presses against cpu more firmly. as good as water cooler is, it's no good if it's not tight enough. for a 300 dollar product you'd think they'd design a block that is better and isn't cooling the bracket assembly instead of the cpu. Unfortunately i can't record a product for being crappy.

The Raystorm blocks are pretty good, actually in the top few percent. I have two, a regular Raystorm and a full copper. I also have an EK supremecy Nickel I'm running right now on a 4790K. But honestly I couldn't see any difference in cooling on an overclocked 4770K between a Raystorm and the Supremcy. I just like the looks of the EK block better, especially since I polished it up so it's clear now.

[boondocks]

Re: Gigabyte Z97 Boards Teazer

Z97X-UD5H
4790K
Highest stable OC: 4.8
uncore/cache: Auto (8-40x)
Vring/cache voltage: 1.2
OC VID: 1.355
OC Vcore: 1.380
Vrin/Vccin: 2.0

24/7 I decided to run 4.7 since I can with 1.272 max vcore which is a little better than my 4770K; however my 4770K will boot at 5.0 with less than 1.4v and the 4790K will not.

Soon as I get Gigabyte to (hopefully) fix my Z87X-UD5H I will stick the 4770K back in it. (I got some bent cpu socket pins )

[wazza300]

Re: Gigabyte Z97 Boards Teazer

To the best of my knowledge you cant adjust LLC on Gigabyte Z87 boards,at least not for the Vcore...you do have one for Vrin though....

theres 4 or 5 levels of llc on z77 low to turbo I think,dunno if its same on haswell but each level will boost cpu by a certain amount to counter vdroop at load

Soon as I get Gigabyte to (hopefully) fix my Z87X-UD5H I will stick the 4770K back in it. (I got some bent cpu socket pins )

surely you can straighten them thar pins?

[MysticalOS]

Re: Gigabyte Z97 Boards Teazer

The Raystorm blocks are pretty good, actually in the top few percent. I have two, a regular Raystorm and a full copper. I also have an EK supremecy Nickel I'm running right now on a 4790K. But honestly I couldn't see any difference in cooling on an overclocked 4770K between a Raystorm and the Supremcy. I just like the looks of the EK block better, especially since I polished it up so it's clear now.

That's what I thought. I just don't get temps i'm seeing. I've redone block 3 times. it's got best contact i can get it to have and good TIM application. Yet the temps are just as high, sometimes even a little higher than the noctua D14 it replaced. The radiator is outside of case with very good airflow. No tube kinks of any kind and flow looks good. Pump is at 5 for maximum flow. The only thing in loop is cpu so no heat off GPU at all. I'm baffled as to why it does such a poor job cooling cpu. I really think it's the 1150 socket bracket though. I don't think it's got enough pressure on cpu compared to noctua which has a smaller sink and doesn't sit on top of bracket too. I'm pretty sure at top right corner the bracket IS taller than the cpu.

My 4790 is not delidded but I am wondering if that'd help. If heat transfer to the cpu lid isn't good then all the heat is trapped in the die and not reaching the water block maybe? I wasn't sure if the 4790ks still have that 0.6 mm gap or whatever it is or not. As for bracket, not sure what to do about that. I really do feel like the block should have thought of that, and not been so huge.

[boondocks]

Re: Gigabyte Z97 Boards Teazer

theres 4 or 5 levels of llc on z77 low to turbo I think,dunno if its same on haswell but each level will boost cpu by a certain amount to counter vdroop at load



surely you can straighten them thar pins?

LLC doesn't work the same with Haswell boards, wazza, it keeps the Vrin up. Haswell cpu's are a whole new ballgame compared to IB.

Tried straightening the pins but broke at least one. Anyway could not get it to boot so I'll just RMA it, GiB is pretty good about fixing this sort of thing. Dummy me forgot Haswell boards the cpu socket cover goes on top the retention mechanism instead of directly on the socket and I mucked the pins up.

That's what I thought. I just don't get temps i'm seeing. I've redone block 3 times. it's got best contact i can get it to have and good TIM application. Yet the temps are just as high, sometimes even a little higher than the noctua D14 it replaced. The radiator is outside of case with very good airflow. No tube kinks of any kind and flow looks good. Pump is at 5 for maximum flow. The only thing in loop is cpu so no heat off GPU at all. I'm baffled as to why it does such a poor job cooling cpu. I really think it's the 1150 socket bracket though. I don't think it's got enough pressure on cpu compared to noctua which has a smaller sink and doesn't sit on top of bracket too. I'm pretty sure at top right corner the bracket IS taller than the cpu.

My 4790 is not delidded but I am wondering if that'd help. If heat transfer to the cpu lid isn't good then all the heat is trapped in the die and not reaching the water block maybe? I wasn't sure if the 4790ks still have that 0.6 mm gap or whatever it is or not. As for bracket, not sure what to do about that. I really do feel like the block should have thought of that, and not been so huge.

Maybe. I've seen some DC cpu's reported as having high temps; it's not completely unheard of.

[MysticalOS]

Re: Gigabyte Z97 Boards Teazer

Maybe. I've seen some DC cpu's reported as having high temps; it's not completely unheard of.

That wouldn't explain the cooler equal to air cooler. But I don't want to derail the topic too much. I may just pull whole liquid system and put noctua back in since it runs at same temp with FAR less maintenance. As for water cooler, I don't know, 300 dollar paperweight?

[boondocks]

Re: Gigabyte Z97 Boards Teazer

That wouldn't explain the cooler equal to air cooler. But I don't want to derail the topic too much. I may just pull whole liquid system and put noctua back in since it runs at same temp with FAR less maintenance. As for water cooler, I don't know, 300 dollar paperweight?

Somethings not right. I've never heard of a Raystorm block not working, and there are literally hundreds of people on OCN with Raystorms. Maybe something else in your loop.
But, your call, do what you need to do.

[MysticalOS]

Re: Gigabyte Z97 Boards Teazer

Loop is literally just pump, radiator, cpu. no kinks, flow is strong. just poor heat transfer from CPU. I really just don't think the block is designed well enough to cover whole cpu effectively because of socket. maybe if it were as small as noctua I could get a tighter fit. I could try sanding/cutting the block down but I'd be worry about piercing it. I just wish the copper was smaller and didn't plant itself over the socket bracket. Maybe I can find just a smaller block to fix the system perhaps.

[boondocks]

Re: Gigabyte Z97 Boards Teazer

Loop is literally just pump, radiator, cpu. no kinks, flow is strong. just poor heat transfer from CPU. I really just don't think the block is designed well enough to cover whole cpu effectively because of socket. maybe if it were as small as noctua I could get a tighter fit. I could try sanding/cutting the block down but I'd be worry about piercing it. I just wish the copper was smaller and didn't plant itself over the socket bracket. Maybe I can find just a smaller block to fix the system perhaps.

It's designed to sit over the retention mechanism...does not hurt a thing since there is no way the water block can sit below the top of the cpu IHS. Having a large surface area on the water block is a plus, not a detraction. I've installed / removed my waterblocks many times with no interference problems from Z68 - Z97.

When I first started water cooling I went from a NH-D14 to an XSPC water setup not as good as yours (old style XSPC water block/ 240mm radiator) and it was a step up from the Noctua.
But like I said, it's yours do what you want with it, I'm not going to try to tell you what to use. Sometimes you're just on your own and have to figure things out.

[MysticalOS]

Re: Gigabyte Z97 Boards Teazer

It's designed to sit over the retention mechanism...does not hurt a thing since there is no way the water block can sit below the top of the cpu IHS.

I think you misunderstand. the socket area is TALLER than cpu. it's not getting ENOUGH pressure on IHS, not too much. if the IHS was taller it wouldn't matter i agree. However, in my case, I'm actually seeing that at least one of the corners, the bracket is taller; PREVENTING the block from applying evenly and effectively to the IHS. a problem I didn't have with noctua because it was smaller and not on top of bracket and thus, could sit on IHS tighter. It's sounding more and more like a possible board problem than water block problem. One that probably has no easy fix. I did notice first couple times I took block off it has a strong engraving of that corner too in the copper from tightening down to get to the cpu, i suppose with enough wear it'll sort itself out too. This is the only thing i can find wrong with the setup. Nothing else is wrong with it that I can find.