• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

The Ryzen 5 3600 Discussion Thread

Just because you have been building them for 20 years doesnt mean that you have been doing it properly.
As you say you dont use stress testers. That in itself shows that after 20 years your still an amature. I doubt any of your overlocks would be able to stand up to anything outside of web browsing.
As said before if your doing it wrong carry on but dont criticise those who are doing it right.
 
Cyber-Mav said:
Just because you have been building them for 20 years doesnt mean that you have been doing it properly.
As you say you dont use stress testers. That in itself shows that after 20 years your still an amature. I doubt any of your overlocks would be able to stand up to anything outside of web browsing.
As said before if your doing it wrong carry on but dont criticise those who are doing it right.
Click to expand...

You're missing the point, because again you're judging a method that TO YOU isn't good enough, my methods work for me, i don't need you to tell me you know better and that i'm wrong, adopting your methods will not improve my results, they just add unnecessary expense.
 
humbug said:
i don't need you to tell me you know better and that i'm wrong.
Click to expand...

on the contrary, you need some serious guidance since you post on here using a lame ass stress test to claim the stability crown then get over excited that your temps dont exceed 53c and praise all over the forums that the new amd cpus run really cool etc etc when the truth is that your lame stress test doesnt pain the real picture.
 
I am finding it imposible to get even a single core to hit 4400mhz with my 3600X. I think my cooler is not helping . I have a corsair H75 hydro. It is hitting 75deg in cinibench.. Any thoughts?
 
Cyber-Mav said:
spoken like someone who has crap cooling and is afraid the cpu will meltdown.
any system i build is tested to 72hours or more of linpak. cpu temps regularly go over 80c on stock cooling with mild oc.

too many clown on here have overlocked systems that fail when even some heavy lifting is presented to them. prime95 is not a good torture test for a cpu. it doesnt load it up enough.
Click to expand...
Can't believe it is 2019 and ppl still burning their (lovely new 7nm) cpus with unrealistic load for unneeded server like stability :D . Don't u think that no1 gonna use their system if it is unstable for every day use? :o
 
Set 100% on the pump rather than auto & now its ok in Cinebench with all cores at 4.4 & fails at 4.75.
Played Metro Exodus & all cores stayed at 4.4 but only at 1080p so the GPU was only running at 57%, so further 4k testing is needed.
I noticed that increasing just 1 core 25Mhz makes most the other cores go haywire & downclock to eg 3.6Ghz. Why is that?
 
Lolik said:
Can't believe it is 2019 and ppl still burning their (lovely new 7nm) cpus with unrealistic load for unneeded server like stability :D . Don't u think that no1 gonna use their system if it is unstable for every day use? :o
Click to expand...

He kind of has a point from browsing the amd threads here recently. Whilst that kind of testing is excessive in my view, getting through a run of cinebench with memory tested for an hour then some games is a bit of a joke. I personally find games occasionally crashing at random intervals annoying, so would rather spend the time initially or just run conservative clocks.

Still, it's not as silly as people saying you don't need bdie then spending silly amounts of time trying to get their cheaper ram to 3600 c16 :D
 
I'm sorry but whether a system is "stable" isn't an opinion. Especially when you have more than one thing in your system running above stock (e.g. CPU, RAM, GPU), if you get a random crash or oddity happen you can never know what caused it if you're not 100% stable on those components. "Was it a software bug or is my CPU not stable?" A workaround is to reset everything to stock and see if the issue re-occurs but that's a huge pain when it might only happen once an evening.

You can have a "stable enough for me" state but personally I'd never accept that. You can't just say "well I never do this kind of workload so it's fine that it crashes during it" - you have no idea what instructions are being used under the hood and where they might crop up elsewhere. You might run an application a few months down the line and see it behave weirdly. How do you know if it's a stability issue if you already know that your CPU is unstable in certain circumstances? How many of us have gotten a system "stable" by doing one or two stress tests, only to find it crashes during gaming?

If you know a component isn't stable in a particular workload then you cannot claim your system is in any way stable. It's not "stable enough" or "stable for what I use it for", it's just unstable. It's difficult to tell because overclocking isn't an exact science but the only way to know your rig is actually stable is doing as many different stress tests as possible. OCCT, linpack, Firestrike, MemTest, etc. Down the line you might find some weird workload that it's not happy with but at least with a wide variety of tests you've done due diligence and can at least confidently call it stable.

Temperatures are a different matter: it doesn't matter if your CPU gets to 90 °C after 24 hours of linpack if you never do anything like that in reality. Just test a worst case scenario for you, e.g. a 2 hour transcode or something.
 
But what if a system is stable enough for your every day needs yet not stable enough to meet someone’s rigorous server like regime? Not everyone is pushing their chips to the absolute limit.

And yes it can be subjective otherwise we would have a standardised suite of stress testing programs acknowledged throughout the community.
 
There's plenty of games that are very susceptible to system instability, so it's not even like you can say 'game stable', as it might be only stable in certain titles.

I know for a fact that Overwatch for example is extremely picky, and if any component isn't solid as a rock it'll fall over very quickly. Division 2 also seems to be really sensitive to unstable OC's.

There's plenty of games I can run at much higher OC's and they'll happily chug along, but fire up either of those two and I know i'll have a CTD or black screen sooner or later.

So i'm very much in the camp of 'no such thing as game stable', it's either stable or it isn't and will subject my PC to a barrage of tests to ensure it's solid. There's nothing more annoying than being mid round or half way through a raid and the game dies.
 
Doobedoo said:
But what if a system is stable enough for your every day needs yet not stable enough to meet someone’s rigorous server like regime? Not everyone is pushing their chips to the absolute limit.

And yes it can be subjective otherwise we would have a standardised suite of stress testing programs acknowledged throughout the community.
Click to expand...
That's the point, there is no way to know what is rigorous and what isn't. Some servers probably don't stress the CPU at all compared to the average game. It's not like there's a linear scale where "80% stable" is enough for games and "100% stable" is enough for servers. You can definitely have a CPU that is stable for 100 hours Prime95 but crashes in a particular game.

The only way to be reasonably confident that your hardware is going to be stable for all the scenarios you might want to use it for, is to test it in as many ways as possible. I wouldn't blame someone for missing out a workload or two because they don't think it's realistic for them but if you do that, you should understand that you cannot say with confidence that your rig is stable. As for people who claim their rig is "stable enough" despite knowing it fails certain stress tests....well they're just lying to themselves.

HeX said:
There's plenty of games that are very susceptible to system instability, so it's not even like you can say 'game stable', as it might be only stable in certain titles.

I know for a fact that Overwatch for example is extremely picky, and if any component isn't solid as a rock it'll fall over very quickly. Division 2 also seems to be really sensitive to unstable OC's.

There's plenty of games I can run at much higher OC's and they'll happily chug along, but fire up either of those two and I know i'll have a CTD or black screen sooner or later.

So i'm very much in the camp of 'no such thing as game stable', it's either stable or it isn't and will subject my PC to a barrage of tests to ensure it's solid. There's nothing more annoying than being mid round or half way through a raid and the game dies.
Click to expand...
Exactly.
 
OK, after spending most of today and yesterday troubleshooting, I can't solve the problem.

If I use any power plan apart from Power Saver, my CPU core voltage stays at 1.4V or higher and my CPU boosts to 4.1Ghz and stays there. That's even while using the AMD Ryzen Balanced plan that comes with the latest chipset drivers from AMD.

Things I have done

I have updated my BIOS.
Disabled Precision boost overdrive.
My memory runs at 3600Mhz, downclocked it to 2666Mhz and 3200Mhz. (There was some Bug with 3600Mhz memory and higher keeping the core voltage high and the boost clock running despite no load. Running lower memory speeds was a solution for some. AMD supposedly fixed it with the latest Agesa version.)
Tried running Vcore voltage at normal instead of Auto(as recommended by AMD)
Reset BIOS to defaults etc.

I have updated windows 10 with the latest chipset driver and have tried the Ryzen balanced plan, still the same.

The only solution for me is to use the power saver plan. It runs at below 1V then when not doing anything.

Not too sure what to try next, I think the problem is with the CPU/Motherboard. When I go into the BIOS the voltage is always between 1.4V and 1.45V. It shouldn't be that high while using the BIOS, it should nearly be at idle Voltages.

I think it's because MSI have only Agesa 1.0.0.3a BIOS for my motherboard and not the Agesa 1.0.0.3ab that has the fix.
 
RavenXXX2 said:
Enable c-states and cool n quiet in bios. Use ryzen master to monitor voltage.
Click to expand...

DragonQ said:
Also try forcibly disabling PBO. My MSI board would always run at boost clocks until I forcibly disabled PBO and enabled CnQ in the BIOS (defaults are broken).
Click to expand...

C-States are enabled, Cool and Quiet is enabled. PBO is disabled.

EDIT: Sorry guys, not sure what happened but only one line of my post was actually posted! Thanks for the Tips, but, as I said above, I had those enabled/disabled and it didn't make any difference. And if I have to manually adjust voltages and offsets to fix it, then something is wrong. So I will just wait until MSI release their next BIOS update.

I am happy with the performance so far despite been hamstrung by running in power save mode. It seems very snappy in windows and games don't seem to be suffering at all. I had a few problems in Window when I first switched over. Was getting a load of Whea-Logger errors and Storeahci 129 errors in the eventviewer. The first was solved by reinstalling the latest Nvidia driver, the second was fixed by doing a windows update.

Hopefully MSI won't be long about releasing an update for my motherboard. I think one of two of the more expensive boards have the 1.0.0.3ab update.
 
Last edited:
Gents if you were to pair with a motherboard buying right now, what would you recommend? I generally spend £100-130 but if decent boards are below this then I am all ears. My last purchase was a Asus TUF, very solid but after couple of years the onboard sound went. I have no clue what brands are now king in the Ryzen era.
 
You must log in or register to reply here.
Back
Top Bottom