Problem with Ryzen 1700 overclock

[Korb]

Hey, guys. I've just recently made my first OC with a Ryzen 1700 and I need some help with some problems I'm having.

My system:

Ryzen 7 1700 withWraith Spire cooler
X370 Taichi (bios 3.20)
Flare X 16Gb (8x2)
Seasonic Focus Plus Gold 750W

OC:

Cpu Frequency and Voltage: Auto > Manual
Cpu Frequency: 3000 > 3700
Am4 Advance Boot training: Auto > Disabled
Voltage Mode: Stable > OC
CPU Core Voltage: Auto > Fixed Mode
Fixed Voltage: 1.21875
CPU LLC: Level 2
VDCCRT SOC LLC: Level 2
C6 Mode: Enabled > Disabled
Blobal C-State Control Auto: > Disabled


Stress Test:

I tried with 1.184v, 1.200 e 1.2065v fixed voltage and even if 5+ cycles of Cinebench went fine, I always had a failed test with Aida64 stability test with CPU, CFU, CACHE and SYSTEM selected after respectively 2, 5 and 9 minutes.

The only stable voltage was 1.21875v. Cinebench was ok after 5+ consecutive runs and so was  Aida64 after 20 minutes.

 

 

The only downside was having high temperatures. Even with fan mode setted on "performance" from BIOS, my max temp was  88.8°C with Aida64. It always stayed there for less than a couple of seconds, then went back to mid-high 70s (76-77°C).

 

Strangely enough, HWMonitor and HWinfo64 read 88.8°C max temp, while Aida64 was on 20°C less 68*C or something. Wasn't this bug only on X chips?
 

I then went on and selected the XMP/DOCP profile from BIOS and RAM was running at 3200.
 

I did all the tests again (Cinebench and Aida64) and the system seemed stable.

 

With this OC, idle temp was between 35°C and 40°c. During heavy gaming it didn't go past 55-60°C.

 

Even if it reached 80°C on an unrealistic stress test, shouldn't this OC be fine since it shows safe temps during normal work?

 

The real problem: when I did a cold boot this morning, the sytem reboot itself 3 times with a beep, while Dr. Debug was showing error F9. It resetted the OC to normal values of 3-3.2Ghz and 2400 for RAM. I l know this is someway RAM related, but what can I do to fix this? I've read that setting VDDCR_SOC at 1.100v may fixed it. Is this true? What can I do?

 

Thanks in advance. 
 

[Brooksie359]

It was probably boot looping to find timings that would get the memory to run at the speed you set in the bios. It likely didn't work so it couldn't post. Lower the ram frequency would be one way or increase ram voltage of it isn't at the highest allowed voltage for the kit. Most allow for 1.35v

[Korb]

53 minutes ago, Brooksie359 said:

It was probably boot looping to find timings that would get the memory to run at the speed you set in the bios. It likely didn't work so it couldn't post. Lower the ram frequency would be one way or increase ram voltage of it isn't at the highest allowed voltage for the kit. Most allow for 1.35v

That's interesting because this happens only when I OC the CPU. If I only select the XMP without overclocking, it will boot normally.

[Brooksie359]

41 minutes ago, Korb said:

That's interesting because this happens only when I OC the CPU. If I only select the XMP without overclocking, it will boot normally.

Honestly that happens more often than you would think. Sometimes overclocking can make the ram overclock unstable and sometimes you need to overclock for the ram overclock to become stable. I had the r7 1700 since launch until I upgraded to the 2700x a couple of weeks ago and had a similar issue with it in the past. I decreased the ram overclock and the issue stopped. 

[Korb]

On 10/7/2018 at 6:25 PM, Brooksie359 said:

Honestly that happens more often than you would think. Sometimes overclocking can make the ram overclock unstable and sometimes you need to overclock for the ram overclock to become stable. I had the r7 1700 since launch until I upgraded to the 2700x a couple of weeks ago and had a similar issue with it in the past. I decreased the ram overclock and the issue stopped. 

I've played around with the settings and I discovered that it wasn't hardware related. What I did was putting Am4 Advance Boot training back to Auto and, with the OC settings above, I was able to cold boot without any error.

[Brooksie359]

50 minutes ago, Korb said:

I've played around with the settings and I discovered that it wasn't hardware related. What I did was putting Am4 Advance Boot training back to Auto and with the OC settings above it let me cold boot without any error.

So it was a ram timing issue in the end like I had suspected.