From Newsgroup: uk.rec.waterways

<div>As you can see from the screenshots below when running cinebench it draws up to 170 watts despite running at a lower clock speed and voltage than the other benchmarks. Is this normal behavior? Is something setup wrong? Not too worried about cause it's only a synthetic benchmark but it also doesn't seem to be typical.</div><div></div><div></div><div></div><div></div><div></div><div>Cinebench R10</div><div></div><div>Download: https://t.co/ZdIrKRneKn </div><div></div><div></div><div>ok. When you are gaming is most used SINGLE core to manage the application.</div><div></div><div>But on benchmarks like cinebench we can choose multicore, so, using multicore the cpu gonna show your true power, more power = more heat. I am testing a long time and i get a conclusion, On 30-| room, water cooler 240mm + dual 140mm intake i can get 85-|, its normal because the cpu is powerfull, we cant compare with a old generation TDP rules. something like 72-|.</div><div></div><div></div><div></div><div>So if you want keep low temps at cost 7% of process power. just disable boost core performace on bios. Using 3dsmax / corona 3d render i can get 60-| maximun temp. Or 75 - 85-| with boost enabled. If you use pc just to play video games, is not necessary change nothing. But pay attention when the multicore games arrive on market. A good deal is force the PWN of the cooler to spin faster x temperature. Do the same to your video card. Set 33% cooler enable at 50-|. anyway, dont hurry, new hardware, new rules. The quality of AIO cooler / water cooler is the material, technology, not size of fans ok.</div><div></div><div></div><div></div><div>4.6Ghz *7051 points = 85-| boost enabled.</div><div></div><div>3.8Ghz *6699 points = 60-| boost disabled. ( no more fan excessive noise, accelerations )</div><div></div><div></div><div></div><div>normal days 20-| room. 72-| with maximum power. Coolermaster nano gel thermal paste.</div><div></div><div>When the environment back to normal, my core boost will be enabled again.</div><div></div><div></div><div></div><div>Ps* Use agesa bios ABBA. good lucky</div><div></div><div></div><div></div><div></div><div>Thank you all guys for helping me. Now it's all solved. It turned out that the problem was the Game Mode in Ryzen Master, an option that force the machine to use only half of cores and threads. I solved all doing a fresh install of WIN10 and reinstalling RM. I noticed that when i unlocked the last 6 core and 12 thread the temps have dropped significantly both in idle (32-38C) and under stress (see cinebench now, the cpu never gets hotter than 67/68C! And i hope this means that i've done a good work with thermal paste and airflow ahah! I hope this thread will be useful to other guys in the future.</div><div></div><div></div><div>I just finished removing all with the utility that misterj linked me above, before trying to install again RM i'm going to do what you said. Once i set bios default i'll try cinebench again then i'll let you know! Thanks guys!</div><div></div><div></div><div></div><div></div><div></div><div></div><div>UPDATE: i set BIOS to default settings and enabled DOCP. I've done another run on cinebench and the results are slightly better but still an half of what they're expected to be. With RM there's no chance, again installation failed. i'm going to open a support ticket for this.</div><div></div><div></div><div>I got my CPU with my new Mobo a few weeks ago, everything is working fine and i have stellar performance in games ( atleast i think so ), i decided to run a cinebench R23 and 3D mark to see if my new cooler Z73 kraken is doing his office, and he is temps are great !</div><div></div><div></div><div>I found some tests on google about the 11900k and cinebenchR23 they are around 16209, with just ABT on, mine with same option is around 15602 but that's fine, the thing that is worrying me is a friend of mine have the same CPU as me and he also have ABT on, and his CPU score on 3dmark is about 14000, mine is only 12500 why is there such difference ?</div><div></div><div></div><div>Don't expect the paste to make much difference, if any. Your temps are maybe a little bit on the high side but not unreasonable for the 5800X. Assuming you don't run cinebench every day, it's not really going to hurt anything to let it run at your temps.</div><div></div><div></div><div>I suppose in good Corsair fashion they dropped PL1 and PL2 so the cooler doesn't struggle (which isn't a bad thing). Have a look at the power limits, see how they are. It's supposed to run cinebench R23 at 253W and complete it before the power limit expires (56 seconds).</div><div></div><div></div><div>I have recently done undervolt and I have passed cinebench to see if it was stable, reducing the negative optimization curve by -20 and the PBO to 110 and I have managed to lower the temperatures in cinebench by more than 10 degrees obtaining the same performance(max temperature package 78 degrees and cores 73 degrees) and in idl package 50 degrees and cores 37 and vrm in cinebench and idl in stable 52 degrees.</div><div></div><div></div><div></div><div>CINEBENCH BERCHMARK WHIT UNDERVOLT:</div><div></div><div> (the dark orange was without undervolt and the light one is with undervolt)</div><div></div><div> dd2b598166</div>
--- Synchronet 3.21d-Linux NewsLink 1.2