Ryzen 5 1600x Low Overclocking Issues

Apr 22, 2018
4
0
10
Hello forums! Sorry for the lengthy post, I just didn't want to leave out any details.

With the second gen Ryzen processors now out, my insecurities about not having the latest and greatest have surfaced :) To counter that and to help make me feel better about myself, I decided to give overclocking my 1600x a shot.

I know that I'm guaranteed nothing over the base 3.60GHz that it comes with stock, however, I had hopes of maybe getting up to 3.80 - 3.90 GHz with a little bit of tweaking. Unfortunately, I'm unable to get above 3.70 no matter what I seem to do.

I'm new to overclocking and realize there's probably something I'm missing. I also realize I may just not have gotten the best silicone on the block, and I'll forever be stuck at 3.70. I know no one likes duplicate questions, so I've been doing a sizable amount of research on similar issues for the past week to try and figure it out on my own, but unfortunately a good amount of it is over my head right now.

I've found some people that said my board (model below) just doesn't have some of the settings that are needed to do more in depth tweaking. I've also read to try adjusting the LLC and the RAM jedec settings, but I'm having difficulty figuring out how to do that.

Anyone have any suggestions? I'm fairly new to this, so I may need things spelled out a little bit :)

Technical information below:

Node 202 case (cooling is not a+, but proc doesn't get higher than 70 under load and idles around 38, and is usually at 45 with normal use)
Cryorig C7 cooler
AMD Ryzen 5 1600x
ASRock Fatal1ty X370 Gaming-ITX/AC - Bios Rev 3.60 (latest as far as I can tell)
RAM: 2x8GB Corsair CMK16GX4m2A2666C16R
Samsung 960 Evo NVMe 250GB SSD

Proc Overclocking Status:
3.60 GHz - runs fine, no issues
3.70 GHz - runs fine, no issues, stress tested successfully
3.75 GHz - boots, crashes on stress test after 90 seconds - temp was only 64C at crash
3.80 GHz - will not boot at all, requires CMOS reset

Voltage is at 1.375 right now. Cannot get above 3.70 GHz stable even by increasing voltage to 1.40 or 1.425

RAM Overclocking Status:
Stable at 1.4v @ 3200MHz

I'm happy to provide pictures of my BIOS settings as well if anyone thinks that will help figure out the available options.

Any help is much appreciated. Thanks!
 
Solution
Cpu-z will be right .

Looks like you got a low leakage chip, which is great for voltage & medium overclocks but not great for high overclocks.
I am by no means an expert but it could very well be, as you said yourself, a case of bad silicon lottery;it can go either way. I have a 1600 and I was more than blessed as it ended up being one of those "rare" 8 core 1600 cpus and its currently running at 3.85@1.275v.

Although your motherboard is not necessarily the best equipped for overclocking I can't see it being the main culprit.
Have you tried running the RAM at stock speeds and then trying to push the CPU?

Hopefully some of the more experienced members will be able to point out the obvious and help you out.
 
I think maybe dropping ram to 2666 would get you that extra 200mhz on the cpu.

In all honesty though 1.375v is about as high as I'd personally go, cutoff for me with ryzen is 1.38v or so.

That c7 isnt really capable of cooling if pushing 1.4v

You'd be looking similar overall performance at 3.8ghz with 2666mhz ram as you would 3.6ghz with 3200mhz ram

Id just stick at 3.7 in a node 202 with that cooler personally.

 
Apr 22, 2018
4
0
10
Hey guys, thanks for the awesome suggestions. It looks like dropping the RAM to stock (1.20v, 2666) allowed me to boot at 3.80 GHz. Anyone know how the RAM clock and CPU clock affect each other and what could explain that? I thought they were pretty separate.

I ran a couple of passmarks without issue but don't have time to stress test it right now with something a little more hardcore like AIDA. When I do, I'll check the real voltage and report back. It is worth noting that my passmark dropped from a little over 14000 on the CPU to about 13700, so that's interesting. MAybe something to do with lesser RAM performance now?

I also agree that I probably don't want to go above 1.375, especially in this case. I'm testing open-case mostly for tinkering and learning purposes and will keep it at 1.375 for daily use. That's hat I got 3.70 to run at stable anyways.

As a side note, I got 3.80 at 1.4v, but didn't even try anything higher or dropping the voltage, that's about all I had time for tonight. I'll do some more testing as soon as I can and report back.
 
Apr 22, 2018
4
0
10
Hey guys, sorry for being absent here. So I found the "issue".

I tried dropping RAM to stock and that didn't seem to help much.

When I switched to manual CPU settings in BIOS, it automatically set the voltage to 1.375v. I had incorrectly assumed that since that was the default, that was the minimum you'd want to set. I did some more googling and read some posts about people overclocking at 1.2-1.25v, so I figured I'd give it a shot. After hours of playing around, I actually got a stable overclock at 3.75 at 1.225v.

I had a lot of success at 3.8 as well, however I noticed that after stress testing (temps only reaching 65-70C), I was unable to reboot immediately. If I tried to reboot, computer wouldn't post. Same with a cold boot. I had to shut down, leave it 10 minutes, and then I could boot. It seemed almost like a "cool down" period, but I find that odd as the CPU didn't reach it's throttling temp of 75C before doing this. I never has OS crashes at 3.80 at 1.225, just post issues after stress testing. During this whole process, I was able to keep my RAM at 3200MHz at 1.2v which was nice as I feel like the RAM is going to give me more results than the CPU overclock for my use case.

I'm happy to keep it at 3.75 because honestly I'm not doing too much with this CPU (I doubt I need any overclock at all), but running nice and cool is my main concern. I am curious though, anyone have thoughts as to what would cause posting issues like that and any potential remedies?

As far as voltages go, CPU-Z shows me around 1.232 and AIDA64 CPUID shows me at roughly .615v. Those numbers are the same whether I'm "idle" with a normal workload, or stress testing with AIDA64. Which number should I believe?

Thanks!
 
@spongebat
Low leakeage chips can be undervolted generally at stock or even small overclocks like yours.
High leakage will overclock well but will require excessive voltages to do so.

Think about the old fx 8 cores, primarily all the same cpu.

Low leakage ended up as the 95w e series, high leakage ended up as the stupid 180-220w 9*** series.

@countmike - cpu-z (the latest one) appears to be rock solid at reading cpu voltage on ryzen chipsets.