• 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.

*** Official Ryzen Owners Thread ***

AMDMatt said:
That's good mate, not golden but decent. I believe mine is stable around 1.37v so not far off at the same clocks.

Voltage is fine up to 1.55v if you have robust water cooling solution. Otherwise try to stay at 1.45v or lower.
Click to expand...

I take it you haven't used linpack for testing using avx?
 
Wow. Guess where that's where the crosshair flexes its muscle - I was getting shut downs at those kinds of voltages with lower clocks - lowered the voltage and could pass.

Some safety feature somewhere kicked in - I wish I could disable it :D
 
eyetrip said:
Wow. Guess where that's where the crosshair flexes its muscle - I was getting shut downs at those kinds of voltages with lower clocks - lowered the voltage and could pass.

Some safety feature somewhere kicked in - I wish I could disable it :D
Click to expand...

Maybe, maybe not.

I found at 4.1Ghz i had to set my DRAM voltage to 1.5v or IBT failed. I'm still learning, but it's more than just CPU Vcore it seems once you get to a certain frequency.
 
I guess a Ryzen owners thread is the best place to ask this although I doubt if many will have this game.

Has anybody gone from a Intel (Haswell onwards) setup to any variety of Ryzen who plays Train Simulator 2017? If so have you gained anything? I have heard people often saying that Ryzen gives a smoother gameplay so has that annoying judder in the game gone? I am actually not expecting much as this is mainly a single threaded game that will use half a extra core or two every now and then (mostly the more modern routes) so IPC is essential for this game. It will probably never be optimised for Ryzen either. Just curious as to how people have found it plays.
 
Don't play that game.

Had i5 4690K @ CPU: 4.9GHz Cache: 4.4GHz RAM: 2400MHz CL11 @ 1T. This was a 24/7 stable OC which I had thrown 48 loops x264 / 4hrs RB stress mode / 17hrs f@h with some 2hrs gaming "back to back" tested when setup it up ~1yr+ ago. Never had an issue in f@h for countless hours after that OC setup or normal usage.

R7 @ 3.8GHz does feel "smoother", it is strangely good "feeling". I have not been able to capture this "feel" in FPS / FrameTime, etc. I even thought is it "placebo" effect of platform change and too many members I spoke to on OCN owners thread get the same feel of "smoothness".

If one self made word describes effect, I would say it's like "FreeSync+" :D .
 
Thanks for the response but if you don't play that game it's irrelevant. I rarely play anything else and my pc is mainly a gaming pc. The game engine is quite old now and mostly uses a single core which is why I asked for peoples input who actually play it.


***Edit*** Didn't mean that to come across as harsh as it sounds but I need input from users of that one game.
 
AMDMatt said:
Maybe, maybe not.

I found at 4.1Ghz i had to set my DRAM voltage to 1.5v or IBT failed. I'm still learning, but it's more than just CPU Vcore it seems once you get to a certain frequency.
Click to expand...

I don't think I have the balls to put that much voltage through memory that would cost me £600 to re-bought :D

I had it up to 1.45v briefly to see if it got me stable at 2666, but no.

Hoping for increased compatibility with my ram in future - I am sure being able to set 2t above 2400mhz would help.

Not sure why that is implemented the way it is currently - seems daft as all hell.
 
Welcome, you wo
eyetrip said:
I don't think I have the balls to put that much voltage through memory that would cost me £600 to re-bought :D

I had it up to 1.45v briefly to see if it got me stable at 2666, but no.

Hoping for increased compatibility with my ram in future - I am sure being able to set 2t above 2400mhz would help.

Not sure why that is implemented the way it is currently - seems daft as all hell.
Click to expand...

I'm sure elmor stated that ddr4 is good for upto 1.8v.
Don't quote me on that I'll try and find his post.
 
pastymuncher said:
I guess a Ryzen owners thread is the best place to ask this although I doubt if many will have this game.

Has anybody gone from a Intel (Haswell onwards) setup to any variety of Ryzen who plays Train Simulator 2017? If so have you gained anything? I have heard people often saying that Ryzen gives a smoother gameplay so has that annoying judder in the game gone? I am actually not expecting much as this is mainly a single threaded game that will use half a extra core or two every now and then (mostly the more modern routes) so IPC is essential for this game. It will probably never be optimised for Ryzen either. Just curious as to how people have found it plays.
Click to expand...

Cannot say I have but I got off the ' Intel gravy train 2017 ' - I did not help you but funny how you see something and it triggers off a thought, saw ryzen and the word train
 
pastymuncher said:
Thanks for the response but if you don't play that game it's irrelevant. I rarely play anything else and my pc is mainly a gaming pc. The game engine is quite old now and mostly uses a single core which is why I asked for peoples input who actually play it.


***Edit*** Didn't mean that to come across as harsh as it sounds but I need input from users of that one game.
Click to expand...

NP :) , no offense taken :) .

Only my opinion, yeah IPC is great on Ryzen but it's not gonna hit the clocks Intel does. PDF in this post on OCN, page 19 has 1700X vs i7 6800K, page 36+ have some more info.
 
On my way to completing 10 runs of IBT at 4.1Ghz. I know this is stable, but just dialing down my CPUSOC voltages. Under 100% load Vcore stays at 1.469.

1bGijMP.jpg
 
I've noticed only by touch mind you that the side of my case is warmer with my 1700 running in it than it was with my 3570k.

Need to upgrade my exhaust fan I think. It's just a generic 2008 exhaust fan.
 
Interestingly, playing games at 3440x1440 I am seeing no FPS difference between 3.6Ghz and 3.9Ghz.

I am most likely GPU bound on every single game. Would anyone else think this?
 
AMDMatt said:
Interesting, thanks. So actually voltage is most likely lower than what i am seeing, and could be around 1.45v or so.
Click to expand...

Current consensus is that. Use CPU Core Voltage (SVI2 TFN) MAX voltage as what CPU is getting, ProbeIt VCORE point as MAX "power plane" voltage in a way. Any "real" under/overshoot on voltages would not be captured on DMM or SW as they are not quick enough, only oscilloscope would.

This is why it gets hard to say to someone don't use LLC, as the user just reads DMM or SW and says "ahh it says this ....", so LLC LVLx is "AOK".

For example my 3.8GHz +162mV OC = ~1.380V (+/- 0.005V) on ProbeIt VCORE point, in OS x264 loading CPU. MAX CPU Core Voltage (SVI2 TFN) = 1.356V, this makes sense as the ProbeIt point is measuring inc "power plane" LLC, all the way from there to socket there "losses" so 0.024V (+/-0.005V) would make sense IMO.
 
Last edited:
gupsterg said:
Current consensus is that. Use CPU Core Voltage (SVI2 TFN) MAX voltage as what CPU is getting, ProbeIt VCORE point as MAX "power plane" voltage in a way. Any "real" under/overshoot on voltages would not be captured on DMM or SW as they are not quick enough, only oscilloscope would.

This is why it gets hard to say to someone don't use LLC, as the user just reads DMM or SW and says "ahh it says this ....", so LLC LVLx is "AOK".

For example my 3.8GHz +162mV OC = ~1.380V (+/- 0.005V) on ProbeIt VCORE point, MAX CPU Core Voltage (SVI2 TFN) = 1.356V, this makes sense as the ProbeIt is measuring inc "power plane" LLC, all the way from there to socket there "losses" so 0.024V (+/-0.005V) would make sense IMO.
Click to expand...

I was using LLC2 in the test above. What do you think the 'true' voltage was, roughly?
 
You must log in or register to reply here.
Back
Top Bottom