Overclock.net banner
121 - 140 of 158 Posts

·
Registered
Joined
·
804 Posts
However, the moment I set the SOC manually, the GFX voltage just follows SOC
On MSI b450 setting GFX voltage do nothing for me, cause its max value actually defined by Vsoc voltage.
 

·
Registered
Joined
·
36 Posts
Found some time to tighten my RAM-timings.

vSoc = Auto
CLDO VDDP = 900Mv
DRAM Voltage= 1.47v . There is stll room to push :) . The DIMM temperature was under 35'C under benching and short stress tests.

2x8GB G.Skill TridentZ 3600Mhz 16 16 16 36 @3800Mhz 16 15 15 15 23 1.47v 1T
Colorfulness Light Product Font Screenshot
 

·
Registered
Joined
·
36 Posts
A little 15 Min. stress test of RAM, at the same settings as above.

2x8GB G.Skill TridentZ Neo 3600Mhz 16 16 16 36 @3800Mhz 16 15 15 15 23 1T 1.47v

Blue Font Line Screenshot Software
 

·
Registered
Joined
·
36 Posts
More time to play wit RAM. Pushing the frequency in this round, and it's scaling pretty good at CL18 .
vSOC= 1.275v
CLDO VDDP = 900mv
vDIMM = 1.48v

The 48.2 ns latency seams to be beating everything out there :love:
2x8GB G.Skill TridentZ Neo 3600Mhz 16 16 16 36 @4600Mhz 18 18 18 18 36 1T 1.48v

Light Screenshot Font Software Technology
 
  • Love
Reactions: mongoled

·
Registered
Joined
·
804 Posts
Nice try. You can do better.
(sorry for being rude)

Font Screenshot Technology Software Number

TM5 stable.
 
  • Love
Reactions: mongoled

·
Registered
Joined
·
36 Posts
Nice try. You can do better.
(sorry for being rude)

View attachment 2529399
TM5 stable.
Of course I can do better. We all can do better, but thta's a pretty good scaling for my sticks.


How stable is your RAM? Do you have a proof of stability too, or just a screen-shot.?
 

·
Registered
Joined
·
2,291 Posts
Of course I can do better. We all can do better, but thta's a pretty good scaling for my sticks.


How stable is your RAM? Do you have a proof of stability too, or just a screen-shot.?
He said TM5 stable

🙂
 

·
Registered
Joined
·
804 Posts
It's my office PC. Since recently I'm not always keen to save TM5 screenshots, because they have little info regarding what testing config was used so one can always take tm5 screen from 3200 cl16 and slap it with MS-paint above whatever settings he wants.
I've run mine with default 1usmus v3 config.
Can do some tests tomorrow.
 

·
Registered
Joined
·
804 Posts
He better show the proof, or it didn't happen
Would it be the proof for you if I made the screen on my home PC in a 5 min and tell you - hey, look, here it is :)
 

·
Registered
Joined
·
36 Posts
It's my office PC. Since recently I'm not always keen to save TM5 screenshots, because they have little info regarding what testing config was used so one can always take tm5 screen from 3200 cl16 and slap it with MS-paint above whatever settings he wants.
I've run mine with default 1usmus v3 config.
Can do some tests tomorrow.
It's always fun to see how far people can push it. We get an idea about the limits of these relatively new chips.

No hurry, post the proof of stability tomorrow, but you can't have my record before that :sneaky:
 

·
Registered
Joined
·
804 Posts
He said TM5 stable
Must admit, that life experience is still a huge factor in quickly
making accurate assessments for situations and people :)
 

·
Registered
Joined
·
36 Posts
Would it be the proof for you if I made the screen on my home PC in a 5 min and tell you - hey, look, here it is :)
I'm not sure if I understand your point. Are you trying to say that people are cheating about proof of stability? Are you dismissing everybody's result, or just your own?
 

·
Registered
Joined
·
804 Posts
I'm not sure if I understand your point.
I just trying to say that sometimes it seems strange when people insists on proofs despite been given verbally assurances, considering what they ask for can hardly prove anything.
 
  • Rep+
Reactions: umea

·
Registered
Joined
·
36 Posts
I just trying to say that sometimes its seems strange when people insists on proofs despite been given verbaly assurances, and despite what they asked for hardly can prove anything.
A Sreen-shot with finished TM5-run is universally accepted as a proof of stability.

It's not easy to fake these high-res images of different windows that cast shadow on each other. It requires a lot of effort, and even then it can be detected. Only a really jalousie person who can't see others success would try to fake it with your suggested methods. ;)
 

·
Registered
Joined
·
804 Posts
Any plans on getting a better motherboard? Though I might pick up a couple 5700gs and see how they perform on my Phantom ITX.
I'd like to, but its an office PC, and I seem to exсeed the limit of investments in my workstation :)
It'd better delay mobo upgrading to the Q1, so as not to irritate my bosses.

Yeah, Phantom itx IIRC 8-layer pcb?
 

·
Registered
Joined
·
804 Posts
It's not easy to fake these high-res images of different windows that cast shadow on each other. It requires a lot of effort, and even then it can be detected
I'll tell you a secret then :)
I almost always compose final screen from a separate screenshots of individual benchmark runs. Solely for the reason of them not to influence each other's results. And, tbh, never bother myself to check, whether someone's screen is fake or not, unless it shows some absurd results.
 

·
Registered
Joined
·
164 Posts
I'd like to, but its an office PC, and I seem to exсeed the limit of investments in my workstation :)
It'd better delay mobo upgrading to the Q1, so as not to irritate my bosses.

Yeah, Phantom itx IIRC 8-layer pcb?
Yup. I also have an Aorus Pro AX B550I that's also an 8 layer that I have laying around that I was meant to sell (person hasn't messaged me in a while though) so might test with both. Don't know. 5700g seems super fun to mess with RAM OCing but next gen APUs next year are supposed to get pretty stupid so might hold off? not sure. Either way, worst case is if I don't end up using it I can sell it in a super small PC to a casual gamer :p
 

·
Registered
Joined
·
36 Posts
I'll tell you a secret then :)
I almost always compose final screen from a separate screenshots of individual benchmark runs.
That's Ok for benching. But that's not the right way for the proof of satbility. You have to put all windows on top of each other and then take one screen-shot. You better do it the right way tomorrow. Otherwise you can't claim my record. :sneaky:

In my proof of stability; I run the AIDA64 benching first, and then the TM5-stablilty test in the same screen.

And, tbh, never bother myself to check, whether someone's screen is fake or not, unless it shows some absurd results.
Oh, yeah? Why have you started the discussion about this topic, and still going then?
 
121 - 140 of 158 Posts
Top