• 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.

SOME RYZEN 7 1700 OVERCLOCKING!

Well, with 3.95 at 1.4V that passed occt linpack, passed an all-nigher encoding - which was the one thing I know failed after passing the likes of aida64 with previous OC's

It's not 4Ghz..... but it'll do if I can rely on it 24/7.

temps didn't get over 60.
 
Not a million miles away now, OCCT is detecting an error rather than just black screening.
302n95w.png


This is 3.9 1.44v LLC2 not completely stable or;
3.8 1.37v no LLC 100% solid.
Not worth it.

This has become a bit of an obsession for me this weekend lol. At least now I know whats happening ; I have a runt!
I would like to know how silicon lottery and 8pack can sell these as 3.9 because clearly not all of them can do it at.
Realbench isn't a true test of stability.
 
Not a million miles away now, OCCT is detecting an error rather than just black screening.
302n95w.png


This is 3.9 1.44v LLC2 not completely stable or;
3.8 1.37v no LLC 100% solid.
Not worth it.

This has become a bit of an obsession for me this weekend lol. At least now I know whats happening ; I have a runt!
I would like to know how silicon lottery and 8pack can sell these as 3.9 because clearly not all of them can do it at.
Realbench isn't a true test of stability.

It seems your cpu is not stable for very heavy linpack which is a shame but all depends on your work load at hand. If you don't run heavy linpack loads and it runs completely fine for everything else you do, I would worry about having the volts so high, you won't have any problems/errors.

Glad you are getting to the bottom of it.
 
It seems your cpu is not stable for very heavy linpack which is a shame but all depends on your work load at hand. If you don't run heavy linpack loads and it runs completely fine for everything else you do, I would worry about having the volts so high, you won't have any problems/errors.

Glad you are getting to the bottom of it.

I'll just be keeping it at 3.8 now. 100mhz with the extra voltage required and the chance of instabilities in the future isn't worth it. If it was linpack stable at 1.44 I'd keep it there as the temps wasn't that high considering. Oh well.
 
Not a million miles away now, OCCT is detecting an error rather than just black screening.
302n95w.png


This is 3.9 1.44v LLC2 not completely stable or;
3.8 1.37v no LLC 100% solid.
Not worth it.

This has become a bit of an obsession for me this weekend lol. At least now I know whats happening ; I have a runt!
I would like to know how silicon lottery and 8pack can sell these as 3.9 because clearly not all of them can do it at.
Realbench isn't a true test of stability.

I would recommend increasing DRAM to 1.5v and CPUSOC to 1.180v manual. That way you can rule out any other voltages from stopping stability. This way you can focus on the minimum amount of CPU Vcore needed to gain stability.

Once you've cracked that, you can begin lowering the other voltages one at a time until you hit instability. This is by far the quickest way to attain stability at 4Ghz-4.1Ghz + I've found.
 
I would recommend increasing DRAM to 1.5v and CPUSOC to 1.180v manual. That way you can rule out any other voltages from stopping stability. This way you can focus on the minimum amount of CPU Vcore needed to gain stability.

Once you've cracked that, you can begin lowering the other voltages one at a time until you hit instability. This is by far the quickest way to attain stability at 4Ghz-4.1Ghz + I've found.

I don't think that this chip can handle anything like 4.0 tbh matt. OCCT fails whether ram is at 3200 or 2133 so I cannot see it being down to ram voltages.
OCCT just passed 2h30m at 3.8 just fine with considerably less voltage than I was using at 3.9 I think this might be a silicon limitation.


2cmt8cz.png


I will try your recommendation and report back, not holding my breath.
 
I just tried OCCT on my 1600 at 3.7ghz at 1.3v. Got to 25mins running at 70c on stock cooler but can't be bothered to let it run any longer on a bank holiday Monday. Will test properly once I have my Kraken mounted and have OCd higher.
 
My R7 1700 - OCCT 3.9Ghz @ 1.316v consumes 90 watts (25 degrees more on air) more than 3.6Ghz @ 1.106v.

These CPUs feast on a large amount of power after stock for fairly small real world gains.

Still keeping it at 3.9Ghz though :D
 
Last edited:
My R7 1700 - OCCT 3.9Ghz @ 1.316v consumes 90 watts (25 degrees more on air) more than 3.6Ghz @ 1.106v.

These CPUs feast on a large amount of power after stock for fairly small real world gains.

Still keeping it at 3.9Ghz though :D

Are you using LLC?
Currently using the settings Matt provided and I'm 30 mins in so far so good at 3.9. Using 1.44v but under load that's dropping to 1.37.
I'd prefer not to use LLC as per raja@asus suggestion over at overclock.net
 
Are you using LLC?
Currently using the settings Matt provided and I'm 30 mins in so far so good at 3.9. Using 1.44v but under load that's dropping to 1.37.
I'd prefer not to use LLC as per raja@asus suggestion over at overclock.net

Mine is all default other than lowering volts, clock and AXMP on RAM to 3200Mhz.

On default there is no Vdroop so many LLC is higher by default.
 
I would recommend increasing DRAM to 1.5v and CPUSOC to 1.180v manual. That way you can rule out any other voltages from stopping stability. This way you can focus on the minimum amount of CPU Vcore needed to gain stability.

Once you've cracked that, you can begin lowering the other voltages one at a time until you hit instability. This is by far the quickest way to attain stability at 4Ghz-4.1Ghz + I've found.

Well what do you know.
1114802.png

Never in a million years would I have said this was DRAM voltage! Forget all that you have learned on intel this is completely different.
Seems I have been doing this wrong all along.
 
My R7 1700 - OCCT 3.9Ghz @ 1.316v consumes 90 watts (25 degrees more on air) more than 3.6Ghz @ 1.106v.

These CPUs feast on a large amount of power after stock for fairly small real world gains.

Still keeping it at 3.9Ghz though :D

Yeh but it is at idle most of the time and even under load, what creates as much stress as things like IBT.

HWINFO us showing 150W peak (CPU+SOC) with IBT using 12GB memory for test. This is at 3800 @ 1.256v.

I think 4000 @ 1.392v was 200W+ and at that point my NH-U14S does begin to struggle.

Games don't take it anywhere near as high.
 
Yeh but it is at idle most of the time and even under load, what creates as much stress as things like IBT.

HWINFO us showing 150W peak (CPU+SOC) with IBT using 12GB memory for test. This is at 3800 @ 1.256v.

I think 4000 @ 1.392v was 200W+ and at that point my NH-U14S does begin to struggle.

Games don't take it anywhere near as high.

I posted back on release power draw with a socket metre for different loads including gaming. Gaming used no where near as much as stress tools. In fact it was very impressive. Posted in the ZEN thread about 8000 pages back lol.
 
Yeh but it is at idle most of the time and even under load, what creates as much stress as things like IBT.

HWINFO us showing 150W peak (CPU+SOC) with IBT using 12GB memory for test. This is at 3800 @ 1.256v.

I think 4000 @ 1.392v was 200W+ and at that point my NH-U14S does begin to struggle.

Games don't take it anywhere near as high.

Here you go sir, https://forums.overclockers.co.uk/posts/30604636
 
Thanks to the guys that have followed this series of events that I have experienced these past few days.
Unfortunately the success I had earlier was short lived I'll break it down here;
1.44v 3.9ghz 1.5v dram = pass
1.44v 4.0ghz 1.5v dram = fail
1.42v 3.9ghz 1.5v dram = fail
1.44v 3.9ghz 1.45v dram = fail

I then went back to the settings that got me to pass the first time, this resulted in a fail. How can a platform be so inconsistent?
Looks like I'm back at verified 100% settings 1.37v 3.8 1.35v dram
 
Back
Top Bottom