Ever cool the motherboard to minimize CPU throttling?

[KingFatty]

So my motherboard has passive heatsinks for the power delivery and other systems.

Currently my CPU sometimes throttles under load while overclocking. I'm still investigating, seems kind of random when it happens.

So I'm wondering if it's an issue with, maybe, something like the motherboard having a temp sensor on the motherboard VRMs and throttling due to heat? Has anyone had throttling issues with their CPU, and then used a fan pointed at the motherboard's passive heatsinks to stop the throttling?

I'd like to work on this, but any tips on a good CPU monitoring tool that will let me see points where the CPU throttles? I think that even when it throttles, it's possible the 'load' would still be at 99%, so I need something that will plot the CPU speed/multiplier/voltage, not just the load.

The cpu is AMD FX-6300, and the mobo is ASUS m4a89gtd_pro_usb3.

 

[Idontcare]

I found that my FX-8350 over-clocked higher and with less voltage after I zip-tied a fan to the passive heatsinks on my AM3+ mobo.

I never saw the 8350 throttle though. Even when I tried to force it to throttle (took it to 83C) it wouldn't, just got to 83C and then the whole thing shutdown like I pulled the plug from the wall.

 

[KingFatty]

Ok good, I will try it. But I simply want to learn more about what is going on in my chip, it's just that I'm constrained in that I'm using a non-supported board with beta BIOS, so it's doesn't support the chip fully. But when I was benchmarking Starcraft 2 for CPU performance, it was interesting how for the same setting, I got different performance by changing the speed of the case fans and cpu fans (I didn't change the frequencies).

Any ideas about tracking the CPU frequency over time, to check for throttling?

 

[Idontcare]

Any ideas about tracking the CPU frequency over time, to check for throttling?

Coretemp. It has a logger feature, configure and enable in the advanced settings section, and it will capture/log both temperature and clockspeed at whatever interval you select.

 

[Exophase]

I never saw the 8350 throttle though. Even when I tried to force it to throttle (took it to 83C) it wouldn't, just got to 83C and then the whole thing shutdown like I pulled the plug from the wall.

So much for what AtenRa was saying about throttling when Tcasemax is exceeded.. not exactly bleeding edge CPU technology either..

 

[KingFatty]

I have heard of motherboards that use a trick. They will flip a temperature flag to read full 255 degrees C (even though the temperature is fine) to force the CPU to cut back its multiplier, thereby forcing throttling of the loads on the power delivery circuitry on the motherboard. I think this is a known trick to prevent exceeding the ability of the motherboard to deliver current, because the FX-8350 can suck down some serious current when overclocked.

Now, knowing about that trick, I think it's reasonable to believe that if the VRMs or other power delivery circuits of a motherboard get too hot, the motherboard can detect this, and then instruct the CPU to throttle itself.

If I had a modern motherboard made for Vishera, I could just look at what is happening by reading up on it. But my situation is keeping me in the dark a bit, because technically I'm not supposed to be using an FX chip on my motherboard, it's just that ASUS was kind enough to release a beta bios to shoehorn the ability to support bulldozer, and it turns out that piledriver also happens to work, so why not?

Anyway, I saw a weird situation where when I cranked my case fans and CPU fan to full maximum, I got better performance without changing any clock settings.

If performance changes due to cooling, surely that can indicate that somewhere there is some kind of throttling?

My CPU temps were low, due to the CPU and case fans being 100%. But, I never paid attention to the motherboard. I'm speculating that by setting the case fans and CPU fans to max, there was a side-effect of also providing extra cooling to the motherboard.

If that's helping my motherboard avoid sending a throttle command to the CPU, then heck yes I'm going to zip tie an auxiliary fan to blow directly on the motherboard. It will just take me a while to set it all up due to other issues taking my time away from computer experimentation...

 

[Idontcare]

I think you have hit the nail firmly on the head.

 

[Rakehellion]

So my motherboard has passive heatsinks for the power delivery and other systems.

Currently my CPU sometimes throttles under load while overclocking. I'm still investigating, seems kind of random when it happens.

So I'm wondering if it's an issue with, maybe, something like the motherboard having a temp sensor on the motherboard VRMs and throttling due to heat? Has anyone had throttling issues with their CPU, and then used a fan pointed at the motherboard's passive heatsinks to stop the throttling?

I'd like to work on this, but any tips on a good CPU monitoring tool that will let me see points where the CPU throttles? I think that even when it throttles, it's possible the 'load' would still be at 99%, so I need something that will plot the CPU speed/multiplier/voltage, not just the load.

The cpu is AMD FX-6300, and the mobo is ASUS m4a89gtd_pro_usb3.

I've known people to cut a giant hole in the side of the case where the motherboard is to lower temps. Works extremely well.

 

[WhoBeDaPlaya]

Definitely. The POS Asrock Pro4 I tried out with a 2500K kept throttling under heavy load due to the VRMs overheating.
Apparently, this is a commonly used technique by ASUS / Asrock.

 

[Magic Carpet]

that my FX-8350 over-clocked higher and with less voltage after I zip-tied a fan to the passive heatsinks on my AM3+ mobo.

I never saw the 8350 throttle though. Even when I tried to force it to throttle (took it to 83C) it wouldn't, just got to 83C and then the whole thing shutdown like I pulled the plug from the wall.

This is a shortcoming of recent AMD chips that I have worked with, they just shutdown, making you lose work. I believe, that mainboards parameters play a role at what temp system shutdown is initiated. I've had the same processor shutting down under different temps, depending on board. e.g. Thuban (85-110c)

And its a must to provide additional cooling to the power circuits to archive any kind of serious overclock.

 

[KingFatty]

Ok got the additional cooling (dual 80mm temp-controlled fans) in place, but I have a question (see below):

I believe I'm fully covering the vertical heatsink, which I believe corresponds to the power delivery.

But see below the CPU cooler, how there is an additional horizontal heatsink that I'm partially covering? I believe it's covered enough, but I could add a baby little fan under there to possibly increase airflow. However, I think the partial coverage blows enough air through and across that heatsink from left to right.

Going to run some benches, see if there is any possibility to get my FX-6300 to 5.0 GHz with all fans maxed out...

 

[69mustang]

I found that my FX-8350 over-clocked higher and with less voltage after I zip-tied a fan to the passive heatsinks on my AM3+ mobo.

I had the exact same experience with my X6 1045T after I glued a fan over the vrm's on my Asus board.

 

[Idontcare]

I had the exact same experience my X6 1045T after I glued a fan over the vrm's on my Asus board.

I never thought to just glue it permanent like. :hmm: I like!

 

[tracerbullet]

There was a time long ago we'd remove the chipset coolers (there were just 2 plastic plugs) and replace them with much larger heatsinks & arctic silver. I used to have a baggie of heatsinks from video cards and old Pentium processors and such. I'm guessing you could still find something like that. If fan cooling works, maybe fan cooling a bigger heatsink would work even better?

 

[KingFatty]

Here's a nice informational post about motherboard VRMs and how you stand to benefit from cooling them.

My motherboard is ASUS, and apparently ASUS uses the CPU throttling technique when VRM temps go too high.

BUT, this article also explains how using non-stock coolers can affect how much airflow goes to the motherboard heatsinks. For example, using a coolermaster hyper 212 or other 'tower' style CPU cooler (or water cooling) can result in better CPU cooling, but worse airflow to the motherboard VRM coolers. So if you have an aftermarket CPU cooler, try to see if your VRM heatsinks are still getting decent airflow - maybe your water cooled CPU is getting throttled because your motherboard VRMs are getting too hot with your FX CPU sucking down all that juice and the lack of airflow from any CPU cooler to blow across the motherboard.

http://www.overclock.net/a/about-vrms-mosfets-motherboard-safety-with-high-tdp-processors

Here is an small excerpt:

VRM Over-Current Protection

Over Current Protection (OCP) is something I have recently been examining. Protection features exist against VRM overheating/overloading depending on motherboard model and brand. I believe it is a crucial feature on motherboards today, because this is the function that will protect your VRMs from a catastrophic failure. This is why I have never seen ASUS boards fail even if people take a lowly 3+1 ASUS boards and try to overclock a Phenom II x6 on it; ASUS boards feature this technology, as it is a part of the PWM controller design.

OCP can work in various ways; one of the ways it works is it downclocks the CPU speed & voltage - via cool'n'quiet or it's own function - if the VRM temperatures are detected as too high (similar to if CPU temps are too high), until they can recuperate and lower in temperature. As a result, it can reduce performance during a full load scenario. It is also how ASUS gets away with rating a few select 3+1 phase AMD motherboards at 125W, though at times the OCP may kick in too often at load even at stock speed/stock cooler and the rating would've been slightly improper for the board (there are few if any 3+1 phase boards ready for 125W processors).

Another common way is a full board shutdown; if MOSFETs are overloaded suddenly to the point where immediate shutdown is needed for protection (i.e. beginning an OCCT run on a 3+1 power phase on a Phenom II x6 OC'ed and at 1.5V), then OCP will kick in and the board will shut down to protect itself. ASRock boards and some Gigabyte boards are known for this.

But some motherboards do not feature any sort of OCP. OCN members and I have found that most recent MSI AMD boards feature NO protection of any sort against VRM failure/over current/over temperature, and this is likely why a majority of the catastrophic failures in the horror stories list are MSI boards. At the moment I and others have been trying to find out which brands/specific motherboards do use over current protection, and we are listing them down for future reference. Once that is done, take it at heart to purchase a board with OCP for your own safety and for the best confidence in overclocking.

 

[john3850]

The POS Asrock Pro3 with my 2500 starts throttling way to early on load so I gave it to the wife for a web use.
I also put at least 2 extra fans on the VRMs on every mb I have owned.

When I ran that 2500k chip at x50 as soon as the temp hit 81c that board would drop the x to 47-48 while using ib.

 

[aigomorla]

oh dear...

sandy-e and the introduction of the z77 got rid of board blocks.

are u telling me the FX boards are gonna require these guys again?

 

[KingFatty]

The article I linked to above mentions how even just a tiny bit of airflow can make a huge difference to the VRM temps, even with a slow/tiny fan. So perhaps it wouldn't be necessary to have full-blown water cooing for the motherboard VRMs, but instead a single tiny fan just to get things moving around really close to the VRM heatsink (if your setup is such that the case fans alone aren't doing enough to compensate for the lack of a stock CPU cooler blowing on those VRM heatsinks).

 

[Rifter]

Cooling the VRM's is always a good idea.

This is why most Artic Cooling heatsinks have curved fins on the bottom 5 or so layers of the heatsink to direct some airflow from the CPU heatsink/fan over the VRM's..

 

[WhoBeDaPlaya]

The POS Asrock Pro3 with my 2500 starts throttling way to early on load so I gave it to the wife for a web use.
I also put at least 2 extra fans on the VRMs on every mb I have owned.

Yup. Didn't seem to have trouble with a 3570K, but it would constantly throttle the CPU multiplier to lower the current draw until the VRM cooled down some. Pointing a black-label Delta at it kept it nice and happy (though not my ears )

Why are we getting less motherboard for the $$$ nowadays :|

 

[Idontcare]

Yup. Didn't seem to have trouble with a 3570K, but it would constantly throttle the CPU multiplier to lower the current draw until the VRM cooled down some. Pointing a black-label Delta at it kept it nice and happy (though not my ears )

Why are we getting less motherboard for the $$$ nowadays :|

It is our own fault. Use a stock cooler and the mobo is cooled by all that down-forced air.

Throw on a 3rd-part HSF and all the air is moved laterally, well above the mobo.

Great for the CPU, not so great for the passive components on the mobo.

 

[Grooveriding]

Yup. Didn't seem to have trouble with a 3570K, but it would constantly throttle the CPU multiplier to lower the current draw until the VRM cooled down some. Pointing a black-label Delta at it kept it nice and happy (though not my ears )

Why are we getting less motherboard for the $$$ nowadays :|

It is my understanding that mobo makers cut costs wherever they can because of having so much competition.

My current board throttled the CPU when the VRMs got too hot as well. I was playing with my CPU when I first got it and was using high LLC, CPU clock would throttle back. Also had my system shut down completely with a burning smell once. Fortunately it lived.

Put a waterblock on the VRMs and it has been fine ever since.