Jump to content

ffbr

Member
  • Posts

    193
  • Joined

  • Last visited

Everything posted by ffbr

  1. @LAR_Systems Pretty cool chrome extension, thanks ! I got my trusty vega56 crunching away, dang that thing is a space heater... Not going crazy this time sick at home (covid negatif) with 30°C in the study, can't really push much more than that. If I find the energy (or the itch grow too much) I might connect on my work account sometimes and then... Good folding you all, and as usual big thanks to @GOTSpectrum and his little helpers for organizing all that !
  2. So I have a bit of an issue, every time I power off my desktop it does not respond to restart command (power button, reset button, power on lan, shorting the power button pin on the motherboard). In order to manage to restart it the following procedure need to be applied: - turn off the power supply - press the power button for 30s (or wait for the mb to empty its residual power) - turn on the power supply while pressing the power button of the computer. Doing a clear CMOS usually allows to restart the computer by pressing the power button however at the next power down the same happen. When the computer is on it works as expected stocked or overclocked. As I am totally at a loss on that one, I am open to any suggestion. History: A bit of history on this problem as it as been ongoing for a while and is still present after some component changes and reinstall It started a bit after adding the CPU to the water cooling loop. When installing the waterblock on the CPU, I had to use quite a bit of pressure as the backplate was not stock and the screws slightly too short. I observed when changing the motherboard that the IHS is slightly scratched on one corner (don't know if it could be relevant) At the time the motherboard was an x470 aorus ultra gaming and it was a different power supply. After reading that it could be a problem with the power supply I changed it with another unit that I knew was in good health (no change) I then proceeded to change the cmos battery, after a bit of hope the problem appeared again As I needed to update my boot drive I also changed it (no change). I also tried to disable fast startup, swap ram with other modules I had around (also tried single channel) Did all bios update and such Tried without any header in the motherboard (sound, usb, fans) As I was really not happy with the VRMs of the motherboard and am planning for a gen 4 high core count I proceeded to change the motherboard, I had a bit of hope but the problem return after maybe 5 reboot without problems) As I was really annoyed I also tried to set up wake on lan, which worked once. What I did observed is that sometimes it would not restart even after turning off the power supply. In that case, removing all usb peripheral allows it to restart using the described procedure) Sometimes letting the computer powered off for long (24h) it will restart on a button press (but not all the time) I have also seen sometimes that the PC still appears as available in some socket processes after power off. All this makes me wonder if the PC does not stay in some kind of active state. Summary of steps taken: - change of motherboard - disable fast startup (windows 10) - change of boot drive - change of RAM - try to use wake on lan - windows reinstall - linux install - bios update - chipset install - CMOS battery change Current components : - Motherboard: X570 tomahawk - CPU: ryzen 5 2600 - GPU: vega 56 (msi aero) - RAM: corsair vengeance LPX 2*8gb - Disk: WD blue SN550 - PSU: seasonic prime 1200 platinum
  3. I personally use virtual box. It's free and I find it fairly easy to learn.
  4. Temperature are getting higher, its getting a bit uncomfortable to have everything running full time. While in winter its nice, allow for "free" heating.
  5. But then in fall/winter, temperature are getting out of hand here. Or not, whatever, it add to the challenge
  6. Dang, they must have recruited their smart fridges too ! (or maybe the cloud is with them)
  7. Might be useful to keep temperature in check for your server room house
  8. Android does it run on smart TV...
  9. Nice, I am now in the world 1%, where is my cake !
  10. For me it keeps saying that they are 8hrs but they complete within 1.5 hours
  11. Well its a server somewhere !
  12. I am at around 50/60pts per 1.5 hour task.
  13. I am now trying to get our monitoring system to understand that those are HS (can"t even ssh into them) so it will restart them. Ok got one back! They are back! Decreasing RAM usage to 95% and crossing fingers.
  14. Greedy Rosetta crashed two of my servers... Might not have been a good idea to allow it 100% of the ram.
  15. Ok ok, pushing two more 32 core servers, once I manage to remote into the instances. I believe they are good old Skylake. Should put about 260 cores on my side, this time it's scrapping under the sofa . I still have around 64 on sprint. It's funny how @leadeater makes me feel all small.
  16. On the rosetta website you can choose the target time under "project preference". The choice are 2, 4 and 8hours with 8 being default. It does not always work but worth a try
  17. And I noticed with a target time of 2hrs the epyc servers are currently pulling around 1h20min per task. ... That is Epic !
  18. I am adding two epyc servers (64cores) (you can say what you want but thats the best server name ever )
  19. So for the sprint, I have a lot of task taking the same amount as the other but that only get 6-7 points instead of 50-70 for others. Do you guys also observe this? and is there a trick to improve the yield?
×