PC Processors

[mpw]

Message was edited by: M W - this answer should not be 'assumed answered' because a solution has not been found.

I'm essentially encountering stuttering in games and some applications, despite my temperatures being fine and low usage of the components. My questions are: is this likely hardware as it is presenting itself on operating systems, over two drives? Also, what other tests can be performed to isolate where the issue may be?

Prior to all of this, I have had difficulties with the system since I purchased the parts, and built it. I have had 3 motherboard replacements due to hard freezing, which appears to have been a PCIe lane problem from PCIe Gen 3 on the board. This is now resolved.

I have also had a problem with my GPU throttling due to heat problems, this was replaced and considered a faulty part.

Description of Problem:

On every setup I have tried, games will encounter some form of short micro stutter. This lasts a short amount of time, and occurs at seemingly unknown* intervals. It can happen in a series of stutters, or not happen for over one hour. Once it begins, it is very noticable and prominent. The stutter will last approximately between 250ms to 1 second. I observe that the frametime will often spike, but not always.

This was tested on both Linux and Windows 10, on separate drives, with fresh installs, over multiple disks. This has been tested under 'diagnostic' conditions, whereby I have installed and updated my drivers, and disabled anything that is not required, nor unrelated to the task at hand.

There are no overclocks, though I have tested with stock speeds at higher voltage. And tested with overclocks. If anything, an overclock will help. Though, I am keeping it removed. This began before any overclocking took place.

I have attempted the following:

* Replaced components

* Reinstalling the OS (and trying Linux, on different HDD's/SSD's, and removing them when other drive is in use)

* Flashing BIOS

* AMD chipset drivers

* Updating Nvidia drivers (and removing old ones with DDU)

* Reseating components

* Removing peripherals and components until the bare essentials

* Resetting CMOS

* Reducing GPU power target

* Checking VRM temps

* Disabling drivers, and keeping the bare esential drivers required to run

* Checking LatencyMon -- sometimes ntoskrnl will spike, sometimes it wont. Sometimes I will have a high ISR latency, sometimes I wont.

* Monitoring voltage

* 3 different monitors, cables and display interfaces (DVI, HDMI and displayport)

* Disabled and uninstalled monitoring software

* Reduced graphics settings in both Nvidia Control Panel and games themselves (remember, it happens in some applications, too) -- VSync off and on

* Using Prime95 for stability at stock settings -- passes all torture tests

* Tried different power plans, with different CPU % settings

* Purchased a PSU tester, and it passes all the tests.

I also disable unnecessary services and processes on both operating systems.

I've replaced the board, the GPU, and almost every other component. I've removed the network card and disabled the wifi and bluetooth drivers. My temperatures are good. To the point where Prime95 will max out around 78C on the test that produces the most heat. The GPU rarely goes above 70C unless in stress testing, and averages at around 61C during load. CPU averages around 58C during load.

There appears to be no throttling, or no obvious sign in monitoring software when this takes place.

The voltages appear to be fine, and hit the target voltages.

I do not know what to monitor anymore. I am kind of at a loss for where to look.

Clearing 'Standby Memory' on Windows appears to help for a short period of time, so I would have assumed disk/ram. Though, yesterday when I 'undervolted' (reduced power target) my GPU, it also helped for a long period of time. This does not always work, though. Sometimes I can remove the stanby memory and it will still happen, just less so. I cannot find the cause.

Component List

CPU: AMD - Ryzen 5 1600 3.2GHz 6-Core Processor Ryzen 5 2600

Motherboard: ASRock - AB350M Pro4 Micro ATX AM4 Motherboard  MSI B450M Mortar

Memory: Corsair - Vengeance LPX 8GB DDR4-2400 Memory x 2 (16GB total) G Skill Flare X 3200 CL14

Storage: Crucial - MX300 275GB 2.5" Solid State Drive Crucial MX500 2.5" Solid State Drive (500GB)

Storage: Seagate - BarraCuda 1TB 3.5" 7200RPM Internal Hard Drive

Video Card: EVGA - GeForce GTX 1060 6GB 6GB SC GAMING Video Card

Case: Thermaltake - Core V21 MicroATX Mini Tower Case

Power Supply: Corsair - TXM Gold 550W 80+ Gold Certified Semi-Modular ATX Power Supply  Corsair RMx 650W 80+

Wireless Network Adapter: Gigabyte - GC-WB867D-I PCI-Express x1 802.11a/b/g/n/ac Wi-Fi Adapter - Not connected

Monitors: Acer - XF240H 24.0" 1920x1080 144Hz Monitor  x 2

Updated component list

[mpw]

I think the threadripper issue may be independent to this.

That being said, I think it would be naive for AMD to say that this isn't a widespread issue. Statistically, it may be on the margin of what could be considered a major issue for AMD, but for the people that have experienced it, they have searched long and far in order to find a fix. It isn't limited to an operating system, and many of us have tried several pieces of hardware in order to rule out what it may be.

At this point, it's not acceptable from neither AMD, nor the motherboard manufacturers to simply say "try a different motherboard" Whilst I will be doing so (I am currently looking for a solid motherboard that have received minimal complaints), I have already spent ridiculously amounts of time, money and effort in order to try and solve this issue.

I receive conflicting advice from AMD and motherboard manufacturers. AMD say to use stock 1809 Windows 10, no modifications, leave all the standard stuff enabled (Indexing, etc) and just update drivers. Which you would assume would be the correct advice. However, motherboard manufacturers are usually giving the opposite advice. I know that ASRock certainly have.

In situations like this, nobody particularly wants to point the finger, they just want the problem to go away -- but I can assure you that the problem is not going away. It is getting worse.

I have provided logs, charts, videos, etc. Yet, all I get from AMD and ASRock is "we don't recognise this issue", followed by "well, a few people said it, but they solved it." Whenever I look online... nobody have solved it. Not a single case of successful resolution from the situation I am describing here. 

Let's look at the facts:

Multiple GPU's, GPU driver updates (from different manufacturers) over different operating systems haven't solved it.

Multiple monitors, single and dual, on the same and different games, with different BIOS versions across different motherboards (different manufacturers) haven't solved it.

The single consistent entity, in my case, is the Ryzen processor. After 12 months, how can anybody say, when being reasonable, that it is not the processor, even when Nvidia said with certainty, that if I replaced the GPU this would still occur because something on the CPU is taking a long, long time to finish (a thread was eating up a lot of time). Despite their advice, I went ahead about purchased an AMD GPU and it still occurred on a fresh install of Windows 10, and Linux Mint. 

Even following AMD Matt's advice and trying AMD/Nvidia combo on a fresh 1809, it still occurs. Even with the latest AGESA updates available. 

I don't know how much more proof AMD need. If I buy an Intel system, and it stops, then that is clearly proof that there is a defect with the processor series. It always seems the Ryzen 5 x600 variants are affected the worst. With the majority of the 7 series able to resolve it, in large part, with BIOS updates and driver updates.

I am bewildered at how simple we are treated, and with disdain. As if we are simpletons unable to come to conclusions after methodical attempts to discover the root cause. 

If this was the legal system, a judge would clearly identify that the issue here is the processor. It's too consistent not to be. 

[mpw]

I have purchased a ASUS PRIME B450M-A that will arrive on Saturday. I will be testing with both the existing Windows 10 installation, and a new install.

I suspect this will not work, and if it does not, I will be then trying a Ryzen 7 2700. If this works, I will be requesting a full refund and compensation from AMD -- paging amdmatt‌

[mittwaffen]

I solved my issue,

My 2600 was stuttering, so i got rid of my video card, tossed in an rx 580. Solved 95% from what it was doing before; the rest is the windows kernal.
The issue I listed before isnt just thread-ripper, its the huge MS delays between threads even on 6/8 core cpus. I think all of the delays add up and if you 
have enough you get the stutter we see. Who knows, it isnt my job to find the cause only to solve it and never look back. 

I had major issues with asus motherboards audio EMI noise even with a soundcard; MSI X470 solved that. 

Still, something to note on QA issues that likely bleed into other areas of power management. 

Avoid ASUS on AMD, We know Asus is best; but when it comes to AMD im sticking with MSI from now on, better by far overall. 
The video card was stable but i think that the video card itself rx 470 has some issues as it was replaced recently aswell under warranty, so some issues are present there with that series. 

So i think it was my video card being a 400 series cause #1, and my motherboard having clear engineering issues #2 and Ryzen 1600 #3 = fixed solution. 

Combination of issues I think, but yeah stay away from the 400 series cards. 

Basically I'd contact ASUS Corporate and complain reference them to this thread. I have corperate contact for AMD (ECR, Lisa's and some of their Engineers -- PM if needed).
RMA for an equivalent 450 series, I think the 350 may have engineering flaws, or design flaws across more then just Asus; however even some Asus X470 and b450 boards have major EMI audio issues. 

I dont appreciate having to deal with issues like this on AMD, it took hundreds of hours to solve. 
However I gave them a chance again on both fronts including the GPU and im happy again. 

If i run into this again with AMD; I will be going full force to corporate w/ letters, emails, phone calls. 
I almost got to that point this time, almost. 

Please AMD work on QA of your vendors or better yet, hire me to catch all of these issues for you.
This was a ridiculous situation no one should ever have to deal with, but thankfully most people would never find the cause of this, since 100 different things can be blamed before finding it -- which is why AMD isnt aware of it. 

[overclock98]

I am using MSI x470 motherboard but its still quite laggy. In a game like CS GO  that's very old it lags on my 2600x and 2060 card

[mpw]

mittwaffen, people seem to be having different mileage with different boards. I've tried a Gigabyte and 4 x ASRock -- no solution. So now I am buying the ASUS. 

I am also going to purchase a 7 series Ryzen.

[mpw]

I have also just ordered a Ryzen 5 2600. I will order a Ryzen 7 1700 for the weekend, and test them all on each motherboard.

My hypothesis is that both the Ryzen 5 1600, and Ryzen 5 2600 with stutter on both of the motherboard, but I suspect that the Ryzen 7 series will not stutter.

I will find out. 

If they do amdmatt‌ (please respond, this is getting beyond a joke) what will AMD be offering up as an apology? This is disgusting service.

[mittwaffen]

Dont go to Matt, go to corporate for this stuff; Matt doesnt have the power to solve this for you. 

Email Lisa directly and get her ECR team.

[mpw]

Lisa most certainly will not even look at my email. 

[mittwaffen]

Im convinced its a software issue.

But who knows, if you email Lisa BTW it goes to her ECR support team to help you out.
Honestly, just send the system into them to reproduce. 

[mpw]

How can it be a software issue when:

I've changed motherboards 9 times in less than a year, changed GPU's and GPU brands back and forth, reinstalled each operating system (Windows and Linux) over different disks? The drivers will be completely different.

The only way this can be software is if it is the AMD Chipset drivers, which Linux doesn't install, and there's an inherent bug in the microcode of the processor. As others claim they don't have this issue, I am willing to bet it's electricity.

If it's not electricity, then yes, it *must* be low-level design and for some reason only a few are presented with the issue.

[shenosuke]

i have R7 1700 with the Same issue... Stutters, Far cry 5, the game that i have more of This stutters.

Things that i already tried /did

CPU: its my second R7 1700 , the first i RMAed because segfault issue , but both showed the stutter issue.

HDD/SSD: i bought 2 New SSDs and 1 HDD, but the issue persisted

MOBO: my first AM4 board (GA AB350 gaming 3) died, so i got refund and buyed a biostar x470gt8 ( that has worst BIOS that i ever seen), the both mobos showed this stutter issue too.

RAM: i have 1 stick g.skill 2800mhz and 1 stick crucial ballistix 2400mhz, i tested both separetly / together, and stutter persisted.

PSU: i tested two Corsair psus(my old cx500 and my New cx650) and unfortunately, the stutters persisted.

GPU: changed my gtx960 2gb to rx580 8gb, both cards showed this stutters.

after ALL of this, i gave up...  Im thinking about send MOBO and CPU for RMA and get the ryzen 3 at launch ( Maybe AMD Solved this issue with this New upcoming generation), or in last case change to Intel, that is too much expensive in my country (Brazil), for u guys have some Idea i5 8400 its like 2x the ryzen 1600 price here.

Sorry for my english

[mpw]

Hi shenosuke,

I have tried new Ryzen 5 2600, and sadly it persisted.

I event tried using a new PCIe network card, incase Realtek GbE was

interrupting with GPU/CPU and hanging.

This did not fix it.

Michael

[overclock98]

I am having the same problem on my PC its been 2 weeks of hell since I built this PC unable to trouble shoot this problem? what do you think should be my next step? take the pc to the shop where I bought it from for a refund?

[mpw]

I would advise testing it in a separate location first. That way you can 100% rule out that it isn't your home/internet.

[overclock98]

https://gyazo.com/dab1577f3f224dbb989e2cc4361953b0

could it be because of this?

[mpw]

The standby memory? That is/was a known issue, however, I keep mine low/restart PC and it still happens.

[mpw]

I have created a list of images of what happens during a stutter. You can clearly identify what is taking place during a stutter: Imgur: The magic of the Internet 

[diasvinic]

I fix my problem, the things that I do: RMA from AMD of my Ryzen 5 1600, they send me another one. I change and the stuter keep happening, so I install a fresh windows 10 HOME and the problem ends. 

My setup now: 

B350 gt3 Biostar

Ryzen 5 1600

Corsair h60

2x4GB ballistix sport 2400mhz

RX 540 Aorus

HD 1TB seagate 

SSD 120GB Sandisk

[diasvinic]

In another forum a guy delet the driver AHCI and the stuter ends if the same setup

[mpw]

Hi, diasvinic‌ -- do you have a link to this forum, please?

Because I wonder if the AHCI driver would be the same between Linux and Windows -- for me, it happens on both operating systems.

[mpw]

I have now tried the ASUS B450M-A, the MSI B450 Mortar and all the previous boards listed. They all stutter with the original Windows install. However, I will proceed to install Windows 1809 from scratch and let everybody know the situation.

My guess? It will continue to stutter. I will try with stock drivers, and then I will advance from there.

[misterj]

mpw, what do you mean by "stock drivers".  You should run ONLY MS supplied AHCI  drivers.  All AMD drivers should be DLed from the AMD Driver Download site.  Enjoy, John.

[mpw]

As in, reinstall Windows 10 on a partition, install Nvidia drivers and

ignore the chipset drivers and test. If it persists, try the Chipset

drivers and test.

[samulelo]

It will continue to stutter for sure i already try with the ahci driver from amd and using the base microsoft ahci driver, today i downloaded cs:go and holy cow if it stutter like unplayable,then i try apex legend that use more the gpu than the cpu is actually playable so is sure a cpu problem but why? and how we can figured out why the cpu don't work properly i have already switch many part even the cpu once.

[mpw]

I've changed the CPU over 5/6 times now. What the hell is this crap...

[mpw]

I am purchasing an Uninterruptible Power Supply. This is truly the last resort. I have spent £1,500+. It arrives tomorrow.

If it is the electricity, then I will sincerely apologise to all AMD, and voluntarily write up a post for them to stick for others displaying the same symptoms and how to potentially diagnose it.

If not, then I expect AMD to address this.

[samulelo]

So does the UPS solved the issue?

[mpw]

The football is on at the moment! I will update you within a few hours.

[mpw]

I tried CSGO but KBoost was enabled. It felt very good. I didn't encounter anything I could really say was 'stutter'. The UPS is plugged in, in a non-optimal way, but appeared to be okay in that sense.

I am now loading League of Legends, and trying with KBoost off. I am monitoring my frametimes. The max for CSGO was 5ms, but with no notable stutter. Which I was very impressed with.

The system overall 'feels' more responsive. Especially during login. I haven't optimised this Windows reinstall yet, so further improvements may still be available.

Leave this with me, I suspect it has helped, but I will do methodical tests shortly and over the course of the next 7 days.

[flayd1]

what the "kboost"? How can off him?

[mpw]

flayd1‌ K-Boost, as I understand it, will set the maximum values for the clocks, and voltages. It should, theoretically, also prevent the changing of P states for the card. Meaning it wont change the clocks.

If you have sufficient cooling, this should not be a problem. Though, be aware, it may increase temperatures. People tend to find performance improvements where they have stuttering or low frames, as the card may be frequently switching between these performance states.

[mpw]

Update 1

The UPS has generally improved performance. I notice a smoother experience from booting, logging in, loading programs, and so forth.

It is entirely likely that my system was fighting dirty electricity, that I was lead to believe (by electrical engineers, and some manufacturers) that dirty electricity wouldn't have been an issue. It's also entirely possible that voltage fluctuations also was causing an issue. I cannot confirm this, as the UPS hasn't actually been caused to switch to battery, but it does have AVR (auto voltage regulation). I notice that sometimes the UPS software sometimes reads 238V from mains, and as high as 244V. 

Whilst the UPS is yet to switch to 'backup' or 'battery' mode, I have experienced generally smoother performance. CS GO and even the main Windows desktop screen feels more responsive, and smoother. League of Legends has improved a lot.

Whilst the above has been stated, I have encountered stuttering both: whilst trying to monitor the frametimes in EVGA Precision XOC, and MSI Afterburner. No *major* stutters, though, as before, with the exception of last night when monitoring CS GO with MSI Afterburner. I noticed a significant frametime spike of around 32ms, which corresponded with a drop in GPU Power. 

That being said, I do not have the UPS setup in the most optimal way until I can get the correct cabling. It is currently connected to a mains 4-way (4 gang way) which isn't supposed to be the best way. So, I would take my experienced with a pinch of salt. There is a distinct improvement, though.

It's also worth mentioning that I do not have KBoost enabled (which could be a separate issue and allegedly cause stuttering through the cycling of P states), and I do have software installed that allegedly in the past caused stuttering (monitoring software such as MSI Afterburner and EVGA Precision XOC). These were supposedly fixed for some systems, but I do not know the reality.

Also worth mentioning is that this is a fresh installation of Windows 1809, and therefore I have not had sufficient time to properly apply Windows 10 optimisations/customisations that may cause Windows 10 to be less-than-optimal for performance. 

I am running stock speeds for the Ryzen 5 2600 (I no longer have the Ryzen 5 2600), and I am now using the MSI B450 Mortar. I am running the 3200 CL14 (14-14-14-34) XMP profile for the G Skill Flare X which may be improved further later on, but I don't want to steer clear away from known-working and tested setups for components until this can be confirmed to be resolved.

I plan to run further tests with HWInfo (as opposed to Afterburner and Precision XOC) and log as many different sensors as possible in order to compare with my previous results when it was stuttering. I will do this in the less-than-optimal setup, and in the optimal setup when the proper cabling arrives.

What I can say, however, is that, gameplay was significantly improved, and smoother. Less stuttering. 

I would like to CC particular people that this would be useful for: mittwaffen‌, amdmatt‌, samulelo‌, misterj‌, overclock98‌ (apologies if I have missed anybody)

If there are any other tests people would like for me to perform for them, please let me know.

I will keep everybody up to date.

[samulelo]

Hi, i would suggest trying other games as in My case i would notice right away if the stutter Is improved as most of the CPU bound game like cs go, fortnite etc. are unplayable, anyway i will bring my PC to test if Is the electricity before buying a ups.

[mpw]

Yes, CS GO was improved. I will try Apex, and Overwatch this evening.

Once again to everyone, I have to reiterate that my UPS is not setup in the most optimal way currently.

[misterj]

mpw, if your UPS really corrects the problem, I would suspect that your house power has glitches/noise on it that the UPS is filtering out.  A possibly cheaper solution, if true, would be a good quality line conditioner.  Good luck and enjoy, John.

[mpw]

Hi John,

Where would the line conditioner be fitted? Between the PC and wall socket

as I have now? If so, this unit was approximately £130. I don't know if

that's a reasonable price or not due to lack of experience in this area.

I haven't fully tested everything yet, so I wont confirm it is solved just

yet. But I will be performing several tests over 7-10 days to confirm.

Thanks,

Michael

[overclock98]

mpw‌ are you able to test the game counter strike global offensive on 1080p settings? I am also using an online UPS with 2kva battery I rarely get any dirty electricity through it

[mpw]

Overclock, I can do this, but I usually play 1280x1024 on medium to high settings, and it was improved. I don't know if changing resolutions would be very methodical for the testings as it changed from the standard test.

What is the rationale?

[overclock98]

For me CSGO/ and PUBG stutter fortnite/destiny2/overwatch/apex legends on low settings dont lag for me

[shenosuke]

mpw can u try Far cry 5 , New Dawn , or Nioh?