Ryzen Stable OC Results

[sbakic]

As title says post your CPU Stable OC results:
To declare your CPU stable you need to pass next tests in next order (lates versions):
1. CinebenchR15 - Benchmark
2. PassMark - Benchmark
3. RealBench - Benchmark
4. RealBench - Stress test (~1h) * select as much RAM as you have
5. AIDA64 - Stress test (~1h)

6. OCCT - CPU Stress test (~1h)  * 64 bit, Large Data Set, Use all Logical Cores
7. OCCT - Linpack Stress test (~1h) * 64 bit, AVX Capable Linpack, Use all Logical Cores
8. prime95 - (Blend stress tests) (~8h) * all workers

Post your CPU model and speed and Vcore at idle (max Vcore)  and load (min Vcore). And your motherboard model and UEFI version.

 *Measure your speed and Vcore with HWiNFO64. For Vcore look at CPU Core Voltage (SVI2 TFN) row or similar.
* Fast way to check if your OC is stable try OCCT - Linpack Stress test for 20 min, a few times. If it is fine run all from above.

* I posted this because I think a lot people don't have stable OC even if they think it is, because they only pass CinebenchR15 a couple of times. This thread should help them to get 100% stable OC.
** And please don't post it if you didn't do all of these tests, or at least OCCT - Linpack Stress test for 20 min. Thanks.

Example:
1700X @4GHz idle: 1.369V - load: 1.331V
ASUS Crosshair VI Hero (Wi-Fi) UEFI: 3008

[deXxterlab97]

Just gonna say that Passmark is a really bad benchmarking software. It claims last I checked a Ryzen 5 1600 is faster than i7 6700 just as is. And also prime95 puts unreal load on your CPU.

Only good CPU bench is Aida and Cinebench. Also @AluminiumTech has a good benchmark software.

[iRileyx]

R5 1600, 3.8ghz, 1.27 vcore(at idle and load)

Gigabyte AB-350 Gaming, bios version F10

 

I'm not prepared to spend 12+ hours doing benchmarks, but it's been stable at this for the last 3 or so months with everything i've thrown at it  

[sbakic]

1 minute ago, deXxterlab97 said:

Just gonna say that Passmark is a really bad benchmarking software. It claims last I checked a Ryzen 5 1600 is faster than i7 6700 just as is. And also prime95 puts unreal load on your CPU.

Only good CPU bench is Aida and Cinebench. Also @AluminiumTech has a good benchmark software.

I don't care about benchamrks scores only stabilty. Point is that I know prime95 and OCCT are unreal load, but there shouldn't be error under their loads.

[sbakic]

1 hour ago, RKRiley said:

R5 1600, 3.8ghz, 1.27 vcore(at idle and load)

Gigabyte AB-350 Gaming, bios version F10

 

I'm not prepared to spend 12+ hours doing benchmarks, but it's been stable at this for the last 3 or so months with everything i've thrown at it  

You want to say your Vdroop is 0V, how is that possible? I think you don't know what Vdroop is.

[iLoveCaptTeemo]

This topic is a joke right?

[deXxterlab97]

2 minutes ago, sbakic said:

I don't care about benchamrks scores only stabilty. Point is that I know prime95 and OCCT are unreal load, but there shouldn't be error under their loads.

If stability then Cinebench aint gonna help much

[sbakic]

Just now, deXxterlab97 said:

If stability then Cinebench aint gonna help much

It will, it is first and the weakest test to check stablity, as you see from the order.

[Jurrunio]

There's a pinned post doing the same thing, just including many older results as well.

 

 

[deXxterlab97]

3 minutes ago, sbakic said:

It will, it is first and the weakest test to check stablity, as you see from the order.

Why run a weak test when you can load aida and test for real stability

[sbakic]

4 minutes ago, deXxterlab97 said:

Why run a weak test when you can load aida and test for real stability

AIDA64 as you can see is not the heaviest stress test and even if your CPU pass it, it will not be really 100% stable.

[Ebony Falcon]

1700x 1.42v 3.95ghz

12 hours Aida 64 

prime 95 12 hours

real bench stress test 4 hours 

Intel burn test 4 hours 

 

max temp was 71 degrees in burn test 

max in other tests was 65

 

generL use and gaming 50 degrees