Jump to content

Dr0xium

Member
  • Posts

    7
  • Joined

  • Last visited

Awards

This user doesn't have any awards

Dr0xium's Achievements

  1. With no iCUE, and killing most background apps. I used HWinfo64 and then plotted the graph of voltages vs clockspeed (observed) of the "master" core in Excel. With Division 2 going, which does some multi-threading, but doesn't utilize all cores, you can see that on all cores, voltage stays sustained above 1.4 and the processor speed remains in boosted range. Roughly 4 cores were utilized to a 'significant' degree while many cores were used lightly. Overall CPU usage never spiked above 84% at any time during the recorded window. Still not 100% comfortable with a processor staying significantly above 1.4V for an extended run period. Yet, CPU never got above 66 Celcius, and most of the time was still under 60 during the run. Total CPU package power in Watts, touched 109w once, all other readings are 80W or below across the entire run (2s polling, not super granular but should be capturing a variety of snapshots, reflecting instantaneous loads).
  2. In the same way as the sarcasm about software applies - bugs exist - my experience with custom builds for more than 20 years (sad to say Im that old) also indicates lots of cpu and mobo combos where auto was not reliable. so you are saying youfeel comfortable seeing voltages during boost up to 1.495 as normal?
  3. Any way to check the Agesa update for Gigabyte Aorus Master - is it “just read the changelog?” Also, looking at the settings, going to mess with iCUE a bit, see if I can reduce or eliminate this behavior through polling options, prioritization, etc.
  4. ICUE from Corsair may have something to do with this. Ive searched a lot of forums this afternoon and someone observed correlation. Ill be careful with causation here but I can say after closing iCUE’s tray application, my voltages and processor speeds became far more dynamic, including entering below 1.2V for the first time. I did switch over to HWInfo64, which did also indicate high voltages and an associated near-constant boost behavior. Once iCUE was closed, the systems’s voltage showed correlation to core boosting, still hitting high voltages for short periods of time, but closer to the dynamic behavior that i would have expected. Again, ill be careful saying “its iCUE”, but this may be an app that together with another monitoring application is essentially a race condition affecting boost/rest states.
  5. Adding another image or two... Taken with phone to avoid improperly influencing with additional “observer effect” processes for capture. Downside is big ugly images.
  6. TL;DR: Brand new Ryzen 3900x build with “auto” settings on an X570 Aorus Master. All BIOS and Driver updates applied. Staying consistently 1.4V+ even though wattages and utilization stay low. What gives - and - will this pose an early wear risk to motherboard or CPU? The longer version... Ok, in most circles, im considered pretty technical. When it comes to modern low level PC stuff, I have to admit that there is a certain level I have to step back and say “im out”. VRM phasing? Great. VCore vs VID voltages and modern boosts affecting and managing them... hrmf. Ok. Lets start there: Ryzen 3900x, X570 Aorus Master, Corsair Dominator. Corsair 850W titanium PSU. Corsair 115i Pro CLC. All in a fractal design mesh define case. Windows 10. So far, so good. I am not over clocking. My intent is to use the parts for some light gaming but more for video editing and im a heavy multi-tasker. Routinely 10+ windows open on a 43” 4k screen with active switching to get stuff done. Im not setting any manual overclock at all. Im seeing boost frequencies but also seeing voltage monitoring from either HWinfo or Ryzen Master reflecting a pretty constant 1.4V+ yes, I've read AMD’s recent reddit thread. Yes, ive applied the new AMD chipset driver in windows 10 1903. I was expecting some voltage management up during boost, but im almost never seeing it drop under light usage. why is core VID voltage often higher than vCore? Is this a long term concern? Why does the motherboard auto target 1.2 and then surge like this to 1.4+? I expect variation but this is a lot more than ive seen in my experience building and running PCs back to the 90s.
  7. For those who were fortunate enough to score some samples of the X570 and 3700 or 3900: I know there were some firmware and other bits over the last month or so - has the experience / performance improved further? Qualitatively, would you say that things are “firm enough” for those who hung back to avoid “day 1” bugs to jump in now?
×