vrm switching frequency

[Minsekt]

is it safe to run it at 800-1000 ? 24/7
as i understand its just cleaner voltage, like a more consistent "wave" of voltage. should not be a problem then i guess.
ive been having this at auto/manual 500 before. if i set this to 800 my max temp drops about 5c. ive also noticed that if this is set to 800, realbench actually shows cpu usage in realtime for me. if set it back to auto or 500 again realbench cpu usage stays at 0 all the time. quite interesting.
vrm temps stay the same. voltages and vccin droop are the same too. max temp on auto is 75c, on 800 its 70c and on 1000 its 67c (all tested with 30min realbench stresstest in exact this order without letting the water cool down, kinda strange) my image editing score is about 25k higher too if this is set to 800 and more.

but my system passed every stresstest ive done and realbench for 10hours even on auto. whys the performance now better tho? shouldnt increase the frequency give me higher temps? if not, whats the downside here?

[Qwinn]

Mine is on auto and I get real time CPU usage in Realbench, so that is kinda odd.

I'm looking forward to seeing any other positive results to testing prompted by this thread.

How long have you been running at 800 vrm sf? Was it set to that when you achieved your awesome oc?

[Minsekt]

it was always been at auto, im still using exact the same settings ive posted in the other thread. the only setting ive changed now is the vrm switching frequency to 1000.
im running it now for 2days at 1000. my cpu temperatur is cooling down faster now to. after a gaming sessions it took like 4-5min for the cpu to go back to 30c idle, now its way faster,like 2min.

the reason i kept trying this is because youve posted that your image editing score is like 200k, even tho my oc is a little higher than yours, ive only got 175k, on 1000khz vrmsf its at 205k ish, weird. (im aware that balanced power option can influence the score, but im always running high performance while benching)

[Arne_Saknussemm]

Raja@ASUS wrote:
Sets the switching frequency of the power FETs supplying processor Vcore. Lower switching frequencies lead to a higher VRM efficiency (small power saving) and lower VRM operating temperatures. Setting a higher switching frequency aids transient response (the recovery of voltage to the applied level after a load condition) – at the expense of heat.

Manual says same thing virtually...thermal efficiency vs transient response

I cannot imagine this has effect on realbench reporting...in my case on auto I get reporting no problem.....or on performance in that way...

Would probably help OCing a bit with voltage stability at cost of heat and maybe (the universe being what it is) longevity of VRM?

If you have good active cooling on VRM maybe temp difference is negligible?

[Minsekt]

hm i dont see any difference in temperatur, but yep, im cooling vrm active with a monoblock, so.. maybe thats why. but ive tried it now several times, if its on auto my image editing score is 25k lower and cpu usage stays at 0

ive seen similar behavor with my ram timings. if i oc my ram to 3000-15-15-15-T2 flash player doesnt work or keeps crashing, but hci memtest works up to 400%. if my timings are 3000-15-17-17-T2 it doesnt crash, hci memtest passes 400% too.. thats so weird.

[SGMRock]

Minsekt wrote:

ive seen similar behavor with my ram timings. if i oc my ram to 3000-15-15-15-T2 flash player doesnt work or keeps crashing, but hci memtest works up to 400%. if my timings are 3000-15-17-17-T2 it doesnt crash, hci memtest passes 400% too.. thats so weird.

That's better than I have got from my RAM, I can get them to boot fine at 15-15-15-35-T2 @3000 but they won't pass memtest, lots of errors. I tried 15-16-16-35-T2 and they went through about 40% of the first pass on memtest before 1 of the 16 instances I had got 1 error so figured that wasn't stable either so I'm running them at the stock XMP 3000 rating that Corsair has of 15-17-17-35-T2 and they are fine. Kind of a bummer was hoping they could perform a bit over spec.

I have run RealBench several times though while OCing stuff and it always shows CPU usage at 100% for me when running.

[Minsekt]

SGMRock wrote:
That's better than I have got from my RAM, I can get them to boot fine at 15-15-15-35-T2 @3000 but they won't pass memtest, lots of errors. I tried 15-16-16-35-T2 and they went through about 40% of the first pass on memtest before 1 of the 16 instances I had got 1 error so figured that wasn't stable either so I'm running them at the stock XMP 3000 rating that Corsair has of 15-17-17-35-T2 and they are fine. Kind of a bummer was hoping they could perform a bit over spec.

I have run RealBench several times though while OCing stuff and it always shows CPU usage at 100% for me when running.

well my ram is stock xmp 2400-15-15-15-T2 , so im more than happy it does 3000 😉

[Arne_Saknussemm]

Minsekt wrote:
well my ram is stock xmp 2400-15-15-15-T2 , so im more than happy it does 3000 😉

That is a nice RAM OC 😄

[Minsekt]

Arne Saknussemm wrote:
That is a nice RAM OC 😄

yes, looks like i won the silicon lottery all around, cpu does overclock very well and my 980ti does 1600mhz too, lol :cool:

[Arne_Saknussemm]

Have a look at this post https://rog.asus.com/forum/showthread.php?44392-Realbench-2-1-cpu-useage-counter&p=377749&viewfull=1...

Where are you running realbench from...desktop? try downloading and extracting a fresh version?