• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Ryzen 5 1600, overclocking results.

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

bos_dc2

Registered
Joined
Jun 8, 2017
Location
Canada
Hi all,

Running a Ryzen 5 1600 on a Asus ROG Strix B350-F & 16GB Corsair Vengeance LED 3000mhz.

Managed to achieve 2933 mhz on the ram and 3.8Ghz with stock voltage on CPU.

Ran Cinebench without crashing and Prime 95 blend test for 5 mins no problem.

Should I be running P95 longer, and if so, how long?

OC RESULTS.jpg
 
If you care about P95 stability, then anything from a couple of hours to several days, depending on how "sure" you want to be. For me four hours has been enough, although I think I've tested up to 8 hours. However, note that you're above safe temperatures already. You should try to keep the CPU below 75 C if you're going to run the Prime95 for prolonged periods. With temps hitting 80 C, I would avoid running P95 for anything more than a few minutes.

Edit: How are you running stock voltages if the Vcore is hitting 1.4 V?
 
Which Prime95? There are test builds of 29.x available with Ryzen support which will use AVX instructions. If you use 28.x or older they don't recognise Ryzen and use old AMD transforms, not sure exactly which instructions but not AVX. Thus they can stress different parts of the CPU. If you run blend, I'd say an hour gives a good sanity check level of confidence, but you can leave it longer for increased confidence. On Intel I find stability can decrease at the same settings on warm days than cooler ones, so if you're on the limit of cooling that could tip you over. Note when it is running bigger FFT sizes, it will stress the ram subsystem also.

Good question asked earlier on voltages. In my experience with a 1600 and 1700, voltages need increasing for stability past around 3.6 or so. Maybe there are "good" samples that can go further, but there is also a possibility the mobo is changing the voltage as you overclock.
 
Those are good questions lol. I don't know if my mobo is automatically adjusting for me. I'll have to take a look.
 
On my wife's Ryzen 1600 I found I needed 1.28 to get it stable at 3.8 using 2400 mhz RAM. Personally, I gravitate more and more to using OCCT as my stress testor. Gives a stress test of equivalent vigor to P95, uses AVX, has a timer, temp and voltage readouts built-in, though for Ryzen the temp readout may not be accurate. Use Ryzen Master for temps.
 
On my wife's Ryzen 1600 I found I needed 1.28 to get it stable at 3.8 using 2400 mhz RAM. Personally, I gravitate more and more to using OCCT as my stress testor. Gives a stress test of equivalent vigor to P95, uses AVX, has a timer, temp and voltage readouts built-in, though for Ryzen the temp readout may not be accurate. Use Ryzen Master for temps.

I feel like a rookie.... Oh yes, I am :-/

I'm sure the voltage was set to Auto. Should I start off with 3.8 GHz & 1.30 voltage?
 
That wouldn't hurt anything. Just watch temps during stress testing. If it's stable at 3.8/1.3 then see if you can lower the voltage until you find the lowest voltage that maintains stability. HWMonitor doen't give accurate temps fro Ryzen. Use Ryzen Master to monitor temsp.
 
I got it run at 3.8ghz and 1.33v

Should I be worried if it gets to about 75c after 5 mins of Prime95 blend test? Passed cinebench no problem.

Is it better if I drop to 3.7ghz and lower voltage? I don't plan on getting a new cooler anytime soon.
 
I got it run at 3.8ghz and 1.33v

Should I be worried if it gets to about 75c after 5 mins of Prime95 blend test? Passed cinebench no problem.

Is it better if I drop to 3.7ghz and lower voltage? I don't plan on getting a new cooler anytime soon.

I have it running now at 3.7ghz and 1.27v. I think I'm happy here
 
Should I be worried if it gets to about 75c after 5 mins of Prime95 blend test?

I wouldn't run Prime95 or OCCT for long periods, but in "normal" use you should be fine. If you want to be sure, run HWMonitor or something similar while you use your computer for everyday stuff and check the max temperature reading periodically.
 
Hey all, I finally have solid numbers and real stability after many hours of tweaking and refining. I have a 3.9Ghz OC on the CPU, with an offset voltage of 0.175 and LLC1 that gives a Vcore of 1.412. 4Ghz just wasn't going to happen without going way over 1.45v it seemed. So, I didn't push. The RAM is now stable with a 2800 OC. DOCP is 2400 and it's dual rank RAM, so not much more I could get out of it without really juicing the RAM with more volts than I would like to. So the timings are 16-18-18-42 CR2. Geardown mode off. Voltage is 1.37.

I streamed Wolcen for over 3 hours last night as well as doing 57.5 minutes of a custom IBT with 12024MB of RAM pictured below. I also slammed the CPU with 13000MB custom P95 for 2 sessions of 17+ hours. Memtest86 for 10+ hours and 4 passes. IBT seems to ferrert out RAM that is close, but not quite stable, hence why I have the CR2 set in the RAM. I might be able to get CR1 stable with a bit more juice, but I'm happy with it now since the RAM was much harder to get stable than the CPU.
Image1.jpg

So, that looks it for now. I don't think I have much room left on temps or voltage to try for higher CPU OC. I might be able to squeeze 39.75 or 39.50 out of it, but that's not really going to see much improvement compared to voltage I might need. As it is 1.412 plus LLC1 shows it hitting 1.45 volts in HWinfo. It's not quite that bad when gaming. Temps are good and voltage isn't as bad. So, looks like I didn't get a 'golden' chip, but it's good enough. I'm happy to get whatever I can out of it. And the learning experience is a blast while doing it. :)

EDIT: Thought I'd post a CB 15 bench. Best score at 3.9Ghz. Oh and notice how badly off the clock is.. that problem is really pissin' me of at this point... I have no idea how to fix that. other than band-aid solutions. And notice my old phenom II scores... 4.0Ghz 401... basically move the decimal two places from your CPU clock speed and you have your CB score with a Phenom II.
Image1.jpg
 
Last edited:
Yes i did Aida64 and theres my R15 score.

How long did you run Aida64? I've found even letting it run overnight doesn't mean your PC is stable. I find that you need to run multiple burn-ins with various programs. Intel Burn Test, is a good one. OCCT Linpacks is a good one as well, and Prime 95 with custom RAM usage. Set it to about 85 to 90% of your RAM and let it run overnight. The linpacks push AVX instructions which are pretty taxing, and if you can run those stable for a couple hours of OCCT, then you are probably stable. Prime 95 is a solid oldie-but-goody. Although OCCT and IBT (Intel Burn Test) seem to ferret out any instability faster (at least that was the case for me).

IBT can give you a quick result if you get errors, or if you get black screened you probably don't have enough core voltage for your OC. I found that with errors and no black screen it was either unstable RAM OC, or needed more SOC voltage. If I black screened I needed more core voltage. Just some tips from what I encountered personally getting my CPU stable @ 3.9Ghz.
 
Aida64 runned for like 2 hours i used to use OCCT and Prime95 on Intel cpu's but not anymore, i played games for hours and hours and copy/past files from hdd to hdd and sdd and all seems to be stable no crashes or errors.

- - - Updated - - -

Thanks for the tips btw.
 
What you do with you own system is the best gauge for "stable" for you.
If ppl run folding/SETI/mining/encoding they want super stable .
For my self I want it folding and Csgo/bf1 stable.
I have had a stable p95/occt system then failed in cs in under 5min
 
For me stable is like 30m to 1hour aida64 and Prime95 like 15minutes to see where temps go and then gaming, i play BF1, Overwatch, Ark Survival, Pubg and if i can game all day long without issues then it's stable i don't need to burn my pc all night long to see if it's stable but that's just me.
 
Back