Generally i don't even bother with long stability test. I do 15 mins in a stress test and if thats good then i use my system as usual and see how it goes. Some games will crash a system that can pass any of the benchmark tests for multiples hours so there really isn't a fool proof test other then using your system imo. Doing long tests just seems like a waste of time to me.
-
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.
*** AMD "Zen" thread (inc AM4/APU discussion) ***
- Thread starter Boomstick777
- Start date
Generally i don't even bother with long stability test. I do 15 mins in a stress test and if thats good then i use my system as usual and see how it goes. Some games will crash a system that can pass any of the benchmark tests for multiples hours so there really isn't a fool proof test other then using your system imo. Doing long tests just seems like a waste of time to me.
Soldato
- Joined
- 5 Sep 2011
- Posts
- 12,876
- Location
- Surrey
Depends on the platform as to what tests are the best methods, the cpu architecture can play a part in what works best and what can pass no issue.
Generally speaking Realbench for 1 to 2 hours will riddle out most instability. The problem is with Ryzen, is that generally that bar needs to be lowered when approaching 4Ghz and expectations need to be realistic.
As for the 1800x, that thing is pretty much good for running stock, and not much else.
Associate
- Joined
- 3 Oct 2014
- Posts
- 1,773
What I thought was my computer stable, Adobe Premiere has just ruined my overclock.
30 mins Realbench - No issue
3 Hours Folding - No issue
2 mins of Premiere encoding a test video with a few effects crashes my system two minutes in.
Time to crank up the voltage.
30 mins Realbench - No issue
3 Hours Folding - No issue
2 mins of Premiere encoding a test video with a few effects crashes my system two minutes in.
Time to crank up the voltage.
Pretty much why i stopped doing long stress test runs. My 4770k at 4.4Ghz was stable and could run any stress test i threw at for multiple hours until I got BF1. In that game it would bluescreen within 5 mins every single time so i dropped down to 4.3 and its been fine since. Stress testing is never going to find every single issue no matter how long you run it.
Associate
- Joined
- 10 Mar 2017
- Posts
- 22
- Location
- United States
Yes, hopefully they are right that a BIOS or microcode update can fix this. Overclocked chips being seemingly unaffected is amusing.
EDIT: According to their source, it will indeed be fixed in a software update.
Yes that's what i wrote. It feels the same. But i will update my statement when tonight try to play the game more than two battles.
I play the game since 2011 (54000 battles) having used the following cpus. 1090T, FX8350 @5Ghz, 4820K @4 and 4.8Ghz, 4930K at 4.5Ghz, 6700K @4.7Ghz and now just default stock settings 1700X.
1090T and FX8350 are at the bottom of the performance. 4820K gave me 50% boost on fps. 4930K was the same perf as the 4820K. then 6700K gave me another 20% boost. (with Nano and Gtx1080)
Using the 1700X feels exactly the same.
Anyone else with Ryzen can tell us their observations with WOT.
One thing that I'm still confused about is performance when streaming while playing a game. If I'm looking to stream and play on a 144hz monitor, considering Ryzen is dropping frames to the 7700K, will the deficit close if using a Ryzen chip? In other words, can the 7700K keep the fps lead while streaming when compared to a Ryzen?
Short answer is no.
Video few pages back that touches on this. 7700k doesn't even compete when it comes to high fps whilst gaming and streaming simultaneously.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
Yeah, I got bored of stress testing like 2 months after I built my first computer in probably late-ish 90s. It's worthless, there is no such thing as stable, and never has been, just stable in the workload being used.
Regardless of if you can't pass say prime95 at 2.5Ghz on a T-bird, but can game with never a crash at 2.8Ghz, or if prime 95 passes at 2.8Ghz but multiple games will crash above 2.5Ghz. Testing with prime 95 or whatever newer programs you might use is worthless unless you bought a system to run those stress tests.
If everything your system runs is fine at 4Ghz on a Zen, but a stress test won't pass above 3.8Ghz, do you refuse to use the extra speed when it's fine in games? I've never paid attention to people who bang on about "this passed 24 hours in such and such", who cares. If it's stable in what I actually use the computer for... it's stable, if it's unstable in some program I never used and thus won't cause my computer to crash... it's still stable.
Caporegime
- Joined
- 9 Nov 2009
- Posts
- 25,353
- Location
- Planet Earth
If you are gaming,a better test is probably to loop 3DMark for a few hours.
Wouldn't you just use the iGPU to stream?
Might depend what game you're streaming too... the lead for 7700K in CSGO is massive (around 200FPS at 1080).
It's an option. Suppose it depends on what software you use to stream. The guy that made the video noticed that despite the high fps and performance the 7700k gives when it came to streaming the game stuttered whilst trying to stream at high quality.
Ryzen offered a much smoother experience for both the streamer and the viewers.
The video in question. I've never streamed but that was my take on it. Talks about it towards the end of the video.
Plays the same as on my 4930K. Would probably notice the difference if I turned V-Sync off but then what would be the point.
Basically what you would expect from a multiplayer game that uses one thread and isn't typically demanding.
Rather than putting up the voltage you should keep it at the unstable settings and try and find a stress test that fails. Do us all a favour.
I would start with some AVX2 Linpack with max memory.
Caporegime
- Joined
- 9 Nov 2009
- Posts
- 25,353
- Location
- Planet Earth
OK,I know I asked this before but what are the solid motherboards,ie,especially B350 ones currently after the latest updates??
That stream section was smoother on the Ryzen but I don't know what he was smoking to be using Fallout 4 as an example for anything.
He spent quite some time saying Fallout was a badly optimised POS which has abnormal reactions to memory speed and the game speed increases with FPS causing further bugs which is why it is deliberately capped at 60fps unless you manually edit files.
So he takes THAT game, unlocks the FPS and runs the stream demo....
He's got Ashes, Doom, Deus Ex and chooses the most broken game. The cynical view is that it creates the right results for the review but I'm thinking it would have been a better review if he didn't stream demo with Fallout.
The lower FPS but smoother output seems to be a theme for Ryzen 7 across the board, which is very good going forward.
What I find fascinating at the moment is how skewed everyone's perception is when it comes to gaming benchmarks.
Both the Ryzen 7 range and the 7700k are getting 100FPS+ across the board on 1080p titles.
Most graphics card start to bottleneck at 1440p so the Ryzen 7 and 7700k reach some level of parity when going above 1080p
Personally the difference from 100 to 144hz for me isn't massive, it feels a bit like diminishing returns after 90 (with something like BF4, 1 etc.)
There's a few exceptions, with CSGO your FPS needs to double your refresh rate otherwise the game feels rough... that's some issue with the Source engine. Not sure how many games are like this.
So even if you're gaming/productivity split is 80/20%, I can't see the reason to get the 7700k. Nothing's worse than using your beast game rig to do some video editing and suddenly feel like you've time travelled to the Intel 486 days.
Last edited:
Man of Honour
- Joined
- 13 Nov 2009
- Posts
- 11,633
- Location
- Northampton
Almost 3 weeks into ownership with the Gigabyte B350 Gaming 3 and 1700. Not a single problem to report as yet, can only run my Ram at 2666mhz rather than 3000mhz but I don't know if that is down to the board or the memory or a combination of both