***The Official 8HR Stability 1700/1700x/1800x Overclocking Thread***

[easyrider]

***The Official 8HR Stability 1700/1700x/1800x Overclocking Thread***

Ok...In Typical OcUK tradition I have decided to create a thread purely for analysing overclock data and findings with Ryzen. I have been reading the other Ryzen threads and its becoming hard to track peoples findings with the chips.

So, my request is to post your clocks running 100% full load for at least 8 hours. If your clock fails before 8 hours then don't post...simple tweak and come back

Now, I know this is not a complete guarantee of stability but its a good place to start from.

Conditions for posting

• P95 small FFTs for CPU/FSB/NB/overall mobo stability
  
• CPU-Z screenshot
• Notepad with username and max clock
• Vcore
• RAM timings (if applicable)
• Temperature
• Type of cooling used e.g. water / air / AIO

Testing the overclock

The following applications/methods are not a requirement, but may help you validate your stability before posting.

• Memtest86+ purely for RAM
• HCI Memtest for RAM/NB
• LinX or Intel Burn Test for CPU & RAM

Warning: the damage that additional heat and volts can cause as a result of overclocking a CPU as far as it will go and then stressing it for many hours is not something inexperienced people should attempt to do. It can very easily end in tears. OcUK will except no responsibility for damaged caused to CPUs, motherboards or any other components as a result. Try this at your own risk.

Mod edit: Can we keep this thread on track and not debate the length of the stability test? Screenshots are encouraged and will help validate your overclock. If you don't have anything to add to the discussion and want to come to ruin the thread then please click the "back" button, in your browser. No more warnings will be given.

 

[easyrider]

AIDA64 isn't compatible with Ryzen yet is it? Thought they said they needed to patch it because they didn't get any Ryzen hardware before launch?

*EDIT*

Yup:

Changed to Prime small FFT

 

[easyrider]

i'll fire up the X370 pro 1700 now

Any luck ?

 

[easyrider]

 

[easyrider]

As you can see it has taken quite a lot of voltage to get even 3.9ghz stable, perhaps of note is the fact that 3.9 is AIDA64 & Realbench stable at 1.34v vcore but small FFT P95 needed the >1.4v vcore.

Now that it's finally completed a run I'm going to start seeing which voltages I can shave and remain 8+ stable, could take a long time but I'll update if I get anything significantly lower or when the new RAM arrives.

Prime really pushes that cpu...Kudos. Good idea get some vcore into the chip, run prime then rinse and repeat pulling back a notch until you find the sweet spot for your chip....

 

[easyrider]

great to see all those threads running on a chip so "relatively" cheap - as easyrider said see how low volts you can go

one thing though - maybe different with AMD platform, but I've never found small prime to fall over with anything other than CPU vcore (unless memory stupidly unstable) - only Blend tests mobo/memory stability

I want to know what the lowest vcore is needed for a 1700 small fft...@4ghz

Then Blend and other stress testing can reinforce ones findings.

 

[easyrider]

The 1700/1700x/1800X are all fundamentally the same chip. Potentially the 1800X might end up overclocking 100-200MHz more than the 1700, but as you note there is so few examples being overclocked we can't really draw any solid conclusions yet.

It's pretty pathetic tbh...

I'm bored of reading " I'm stable @ 3.9ghz at 1.3v with a picture of handbrake or realbench running for ten minutes...

I'm getting my data from another forum now...as the comedy gold posts in the cpu section are quite frankly ludicrous....it's like the cpu section as been invaded by a school trip

The blind leading the blind...

I got my 1700 to 3.9 stable with 1.3v but i don't really want to push it much higher than that. Got the memory to 2666 stable as well but it won't boot at all if I try 2933. Overall though I'm happy with it and after a few bios updates I'm sure i'll be able to push the memory higher.

 

[easyrider]

Just to point out, passing small FFT's does not mean memory OC is stable.
Needs to be on blended for full system check.

The whole point of small FFT is to test what vcore is needed for it to pass...

Once this is done run Blend etc...

 

[easyrider]

An arbitrary level picked by OP.

I haven't seen any research done as to how many hours suceasfully tests stability.

Xtreme system forums usually run 48 hour tests...but they know what they are doing...

I chose 8 as it should be easy enough and allow results of what vcore is needed to pass...

Shame the results are so thin on the ground

 

[easyrider]

Many high clockers ignore Prime95 due to unrealistic AVX loads.

I'm not arguing with you post your results or don't...

 

[easyrider]

To pass a sustained SmallFFT run is likely to require a Vcore and subsequent heat output that is beyond what most other CPU stress tests generate so it's a good starter for absolute stability, throw in LargeFFT for testing the IMC/RAM as well and it's a pretty solid base for stability testing. 8 hours is about a full test cycle of Blend (which is a mix of small and large FFT tests).

Where is my like button

 

[easyrider]

Kinda what I was getting at, You may as well run the blended for whatever length of time.

Also, 8 does seem a lot. 2 or so?. I'd either have to leave it on while at work(Not happening) or in evening at home thus not been able to us it.

I run overnight...I have also run prime and posted on ocuk and surfed the net at the same time.

Stop looking for excuses...If you dont want to post then fine...refrain from posting in this thread...

 

[easyrider]

https://en.m.wikipedia.org/wiki/Prime95#Use_for_stress_testing

In the overclocking community, a rule of thumb is often used to determine how long to run Prime95: test the CPU (8 kB FFT) for 10 hours and the memory (4096 kB FFT) for 10 hours, and if the system passes, there is a high chance that it is stable. Twenty-four hours of testing is recommended to be sure, as errors may show up after 16 or more hours of testing (compared to, say, just four hours of testing).[7] Moreover, a large proportion of system overclockers and enthusiasts favor Prime95 over other benchmarking suites because Prime95 pushes the CPU's floating point units extremely hard, causing the CPU to become extremely hot. In addition, Prime95 stresses a computer far more than the majority of software-based torture suites. The nature of this is because the operating system usually shuts down the floating-point unit when unused by other programs, whereas Prime95 is well-optimized to continuously and effectively thread the FPU, causing it to be deeply pipelined, thereby generating significantly more heat because of elevated power consumption under the massive workload conditions. In CPUs which are not adequately cooled, errors are likely to occur. Prime95 also constantly accesses main memory at up to 60 MB per second. This constant activity will detect memory problems that other programs will not.

All this discussion on prime is avoiding the point..either run the tests or don't but please can we stop arguing about the nature of the test...if you don't want to run it..then please don't post and derail this thread any further...

I run small FFT for 12 hours then I run blend...been doing it for years.