FX-8350 goes from 4520mhz to 1406mhz when heavy gaming

CPU may not need to stay locked in at full core speed if the demand isn't actually there. You can check to see if the core behavior is acting normal or not by running Prime95 version 26.6, which is one of few utilities that actually enforce a steady state load at 100% usage, and with HWinfo open see if core usage sticks at 100% for all cores or if there is still fluctuation.

Normally, you don't want the CPU to stay at full peak core speed/voltage, unless it's ACTUALLY needed to be, so fluctuations under normal conditions would indicate that core speed/voltage DOES drop when that core does not need to be at full speed. This is helpful in saving power but more importantly, in keeping core temps under control. Cores that never get the chance to "relax" tend to run much hotter and contribute to higher package temps.

Still, with "performance" enabled as the profile it SHOULD be min 100% and max 100% UNLESS the profile has been modified. I usually recommend setting the profile to performance and then going in and changing the minimum value under processor power management in the advanced profile settings area to 5%/100%. Normally, if there is ANY need for the core to be at 100%, it WILL be. If it is dropping down from 100% when it SHOULD be under a full load, then of course something is fubared.

I'd run Prime and see if the cores stay at 100%. Also, go in and see WHAT the min and max are set to in the advanced settings for the performance power profile. It may be that the min setting is not 100%. If you CHANGE it to 100%, expect to see higher overall core and package temps afterwards.

HWmonitor is known to be almost laughably inaccurate with many chipsets, as is Open Hardware monitor, AI suite and most other bundled monitoring utilities. HWinfo is far more accurate and has much broader coverage for a wider selection of chipsets. I'd ditch HWmonitor, install HWinfo and run "sensors only".

It actually SOUNDS exactly like throttling to me. Try Prime, but also try HWinfo. Scroll to the CPU core values and keep an eye on the core clocks, core temps and package temp. Also, HWinfo actually even has listings for whether or not thermal throttling is occuring, down below the core clock thermal values, so that might offer a little more insight as well, but mostly you will be able to tell simply by looking at the core values.

Also, it's kind of important to VERIFY what the processor power management advanced settings are set to in the advanced profile settings area of power management. Without knowing that, you don't even know if the behavior is contrary to the settings or not. Could be entirely normal based on what is configured in those advanced settings.

I COMPLETELY don't understand what you are talking about, no offense. None of those numbers make any sense in the scheme of what I thought we were trying to look at.

So, try this. Open Hwinfo and scroll down to the CPU core specifications/sensors data. Expand the HWinfo window so it is as tall as you can get it, from the top of the screen to the bottom. Next, open Prime95 v26.6 and run Small FFT. Not Blend or Large FFT, only Small FFT.

Take a screenshot after about five minutes. Post the screenshot here. Otherwise, I have no idea what you are referring to when you say 20.50, 21.50, 7, etc.

In order to help you, it's often necessary to SEE what's going on, in the event one of us can pick something out that seems out of place, or other indicators that just can't be communicated via a text only post. In these cases, posting an image of the HWinfo sensors or something else can be extremely helpful. Here's how:

*How to post images in Tom's hardware forums



Run HWinfo and look at system voltages and other sensor readings.

Monitoring temperatures, core speeds, voltages, clock ratios and other reported sensor data can often help to pick out an issue right off the bat. HWinfo is a good way to get that data and in my experience tends to be more accurate than some of the other utilities available. CPU-Z, GPU-Z and Core Temp all have their uses but HWinfo tends to have it all laid out in a more convenient fashion so you can usually see what one sensor is reporting while looking at another instead of having to flip through various tabs that have specific groupings.

After installation, run the utility and when asked, choose "sensors only". The other window options have some use but in most cases everything you need will be located in the sensors window. If you're taking screenshots to post for troubleshooting, it will most likely require taking three screenshots and scrolling down the sensors window between screenshots in order to capture them all.

*Download HWinfo


For temperature monitoring only, I feel Core Temp is the most accurate and also offers a quick visual reference for core speed, load and CPU voltage:

*Download Core Temp

When it comes to temperature issues, taking care of the basics first might save everybody involved a lot of time and frustration. Check the CPU fan heatsink for dust accumulation and blow or vacuum out as necessary. Other areas that may benefit from a cleaning include fans, power supply internals, storage and optical drives, the motherboard surfaces and RAM. Keeping the inside of your rig clean is a high priority and should be done on a regular basis.

Prime95 v26.6 is THE primarily accepted way to do the majority of baseline stability and thermal limit testing running the Small FFT option.

Prime95 version 26.6: http://windows-downloads-center.blogspot.com/2011/04/prime95-266.html


Further, you can find extensive information regarding the Intel CPU architectures and specifications at the following link which is a somewhat definitive guide on that subject. The information below is taken directly from conversations with Computronix who is also the author of the Intel temperature guide, found here:

http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html

For AMD systems, specifically Zen/Ryzen, this should offer similar albeit not nearly as detailed information on that architecture.

http://www.overclockers.com/amd-ryzen-overclocking-guide/

This is probably about the most referred to overclocking guide around, and it's principles can be applied to a variety of generations and platforms.

http://www.overclock.net/t/91/ultimate-overclocking-guide

This pretty well sums things up and is equally relevant whether working with an Intel or an AMD system.

I can think of several reasons why x264 encoding or AVX / AVX2 / FMA3 apps won't work as a unilateral metric for thermal testing.

(1) A steady-state workload gives steady-state temperatures; encoding does not.

(2) Simplicity in methodology; most users would find encoding apps unfamiliar and cumbersome to accomplish a simple task.

(3) Most users such as gamers never run any apps which use AVX or FMA, so adaptive or manual voltage aside, it makes no sense to downgrade your overclock to accommodate those loads and temps unless you KNOW you will be making significant use of AVX/FMA/AVX2.

(4) Standardization; Prime95 has been around since 1996; many users are familiar with it. It is TRIED and TRUE.

For the minority of users who routinely run AVX/FMA apps, then P95 v28.5 or later can be useful tweaking BIOS for thermal and stability testing. For others, it is not recommended.


regardless of architecture Prime95 v26.6 works equally well across ALL platforms. Steady-state is the key. How can anyone extrapolate accurate Core temperatures from workloads that fluctuate like a bad day on the Stock Market?

I'm aware of 5 utilities with steady-state workloads. In order of load level they are:

(1) Prime95 v26.6 - Small FFT's (Important. NOT Blend or Large FFT)
(2) HeavyLoad - Stress CPU
(3) FurMark - CPU Burner
(4) Intel Processor Diagnostic Tool - CPU Load
(5) AIDA64 - Tools - System Stability Test - Stress CPU

AIDA64's Stress CPU fails to load any overclocked / overvolted CPU to get anywhere TDP, and is therefore useless, except for giving naive users a sense of false security because their temps are so low.

HeavyLoad is the closest alternative. Temps and watts are within 3% of Small FFT's.

Click to expand...

Yes, definitely possible too.

And now that I think about it, with it being at over 4500mhz it would seem that there MUST be some overclocking going on since the FX-8350 is a 4Ghz stock to 4.2Ghz turbo processor. Shouldn't ever run above 4200mhz unless it's overclocked and then we could clearly be into a voltage or cooling issue. That would even further make it likely that the problem is due to VRM or CPU core throttling. As mentioned also, knowing some specs and probably a LITTLE more information would be helpful to determine one way or the other.

Im guessing some cheap asrock or msi 970 board , or one of the 4 phase gigabyte/asus ones.

Mention of an AIO & max cpu temps of 50c & yet still seeing multiplier drops to 7 (1400mhz) is a surefire sign.

Seen it a million times before.

Can't argue with that, especially on the FX8 and FX9 chips. Those barely run right with stock settings on the cheaper boards.

A lot of the boards will even mention that you need to add cooling to the VRMs if you are using some tower coolers and AIO/water coolers. They need air movement from the fan to help cool the VRMs, and AIO coolers don't do that.

No arguments there either. 100% agree.