Undervolting 4790k and "uncore" keeping all power saving stuff on

[flexy]

The fact this is in a tiny m-ITX case with a tiny m-ITX mobo with way less airflow over the VRMs isn't helping any either. And where is the backplate for the cooler? That fat chunk of metal at the back of the mobo that supports the weight?

It doesn't have a backplate, as bizarre as it is. (On their forum, someone "recommended" that this cooler can also be mounted w/o a back-plate, with those bolts.) TBH this thing looks more than suspicious to me, I would never mount a cooler without backplate UNLESS it's a VERY light cooler....but then why would I want to mount some light, dinky cooler in the first place? Seriously he should just get an EVO or something.

 

[EliteRetard]

So basically this is a custom/alternative mounting method? (Similar what I did with mine where I use the EVO212 clamp for the Thermalright Ultra Extreme).

The problem here (I THINK) is that it's possible that with this custom mounting method you can easily introduce stress, bending etc. eg. by over-tightening the screws etc...and the lack of a backplate doesn't help either, it would make the problem worse. The problem won't be apparent with the original "plastic clip" thingies because with them you are never able to introduce that big a force that it would bend anything as you can with the bolts.

So...as a first test you can try use the plastic clips again (because less pressure)..to see whether it also stresses the board and see whether it gives better temps.

Or..re-mount as you did before, but do NOT overtighten the screws at all, in fact try a very careful mount where the HSF sits "just right" but not more.

Other alternative...look for an option to mount (or an entire new cooler for that matter) with a backplate as well. EVO 212 is rather cheap. It's not "high end" but at least it could solve the problems with bending.

I would also recommend another method of applying TIM (not sure about Ceramique, anyone and their mama uses MX-4)..and then not spreading but a tiny blob or a tiny strip maybe 3/4 of an inch down the die like this:

***

This MOBO has no such brace and the clamp warps the board when closed on the CPU.

If it's really that the CLAMP from the socket warps the board, something is off with the board. A backplate may also be the solution here, but I wouldn't say it's normal that closing the socket would warp the board, not at all.

It's not my alternative mounting bolts that caused the warping...I stacked washers to prevent over tightening the bolts. It was the stock MOBO CPU clamp that was a big problem. Now that I've mangled the crap out of the stupid CPU clamp I'm getting better temps. ~80c gaming with the weaker Scythe cooler and case closed vs 90c with a much bigger cooler and open case.

I feel like my method of applying ceramique is the best option, it's extremely thick and the only way to ensure a thin even coat is to do it yourself. I haven't had this issue with other paste.

I've realized how terrible the Scythe cooler is though, it has serious design flaws. One is the flimsy mounting brackets and lack of backplate. Then it uses a baseplate instead of direct touch (baseplate is fairly even and smooth though). The problem is a huge void where the heatpipes go through and they barely contact the baseplate. In fact very little of the heatsink actually touches the baseplate. I've edited a picture to illustrate:

Look at the highlighted red area, that is the void for the heatpipes. But the area highlighted in white is empty. They barely pinched the hetpipes at all, just enough to sorta hold them in place. Since they are round/uneven however, they make very little contact with the baseplate. There is no solder or anything connecting these parts. The only real contact to the baseplate is the aluminum fin stack and it only makes contact with a thin strip on each side highlighted by the green arrows.

I've since pumped the void full of ceramique but I'm pretty sure this is why I'm still getting sub par temps. If I could fit a bigger better cooler I would...but as far as I can tell this is the only option that will fit (100 x 100mm and 75mm tall). I had a thread asking for help and nobody could find a better option.

Not sure 80c gaming is good enough, I still think I'd like to try to undervolt at load to help reduce these temps. I still don't know how to do that without screwing up the idle voltages etc.

 

[coercitiv]

Taking off Vcore FOR UNDER LOAD will likely backfire since you're just asking for instability, assuming that your CPU is at stock and stock voltage. It's not that Intel deliberately over-volts them.

Just as some people do mild overclocks on stock voltage and have stable systems, so can others undervolt on stock speeds and have stable systems. However, the maximum undervolt value gets smaller with every generation, as technology enables chip manufacturers to use smaller tolerances when setting default voltage.

For example, my mobile 4c Haswell chip is undervolted by 50mV. It shaves off a few degrees while being a conservative undervolt (CPU is stable in Prime95 at -90mV).

Undervolting is done via software using Intel XTU. It is set using the Dynamic CPU Voltage Offset value, so it affects all frequency states. The OP may want to check this software out, as it enables easy undervolting as well as setting other limits for selected processors (Turbo Boost Max Power for example).

 

[Cerb]

The fact this is in a tiny m-ITX case with a tiny m-ITX mobo with way less airflow over the VRMs isn't helping any either.

The case and HSF should be providing tons of airflow for the VRMs.

And where is the backplate for the cooler? That fat chunk of metal at the back of the mobo that supports the weight?

That is a good question. If just to be sure to evenly distribute pressure, a backplate should be used with bolts, else just use the push-pins. An HSF like the Shuriken shouldn't present enough torque when moving the PC to be a problem for the push-pins.

 

[Cerb]

Then it uses a baseplate instead of direct touch (baseplate is fairly even and smooth though). The problem is a huge void where the heatpipes go through and they barely contact the baseplate. In fact very little of the heatsink actually touches the baseplate. I've edited a picture to illustrate:

Look at the highlighted red area, that is the void for the heatpipes. But the area highlighted in white is empty. They barely pinched the hetpipes at all, just enough to sorta hold them in place. Since they are round/uneven however, they make very little contact with the baseplate. There is no solder or anything connecting these parts. The only real contact to the baseplate is the aluminum fin stack and it only makes contact with a thin strip on each side highlighted by the green arrows.

There should be solder connecting the heatpipes to the base plate, and the base fins to the base plate. The plate itself should be slightly convex, to better mate with the IHS of the CPU. One of the highest-performing heatsinks out there, the NH-D15, for instance, uses a similar base plate going to the heatpipes. Direct contact is not a bad thing, but being superior is marketing, not reality. Those gaps exist in other coolers, as well, and are normal. Filling them in wouldn't do good. The contact between the CPU IHS and base plate of the heatsink, and that base plate to the heatpipes, is what matters, and that should be soldered, crimped, or both. The fins are attached directly to the base plate, like the heatpipes. IoW, it's not terrible at all.

If I understand correctly, you have multicore enhancement on. If so, 80C while gaming is not bad. If you turn it off, do temps go down much? With it off, you aught to see something like Prim95 jack the temps up ASAP, possibly into throttling, then with the long-term power limit kicking in, fall back a bit, as it pulls a steady 88W.

If you set the voltage a bit lower than it's showing, with a high LLC, you can probably get it a bit cooler under load, without reducing idle voltage. You may have to fiddle with the exact settings, though, since it still can go over your set voltage in some cases. A small offset undervolt might suffice, too, however.