LTT Official Folding Month 2021!!!

[leadeater]

@GorgonFixed it by messing around in the BIOS, couldn't say which fixed it as I changed a bunch lol. Will figure it out later just want to get the damn thing going

[Elektra57]

I am freaking out. I can't find my name under the team stats. Help.  I have been folding non stop. I have the team number correct and it says my WU's are going to LTT. What am I doing wrong?

Edited November 27, 2021 by Elektra57
Added picture

[leadeater]

7 minutes ago, Elektra57 said:

I am freaking out. I can't find my name under the team stats. Help.  I have been folding non stop. I have the team number correct and it says my WU's are going to LTT. What am I doing wrong?

You are appearing in the LTT team stats

 

"Elektra57","74962988","2818","223518"

 

And you are also in the LTT Folding Month stats as well.

 

Where are you looking and what are you not seeing?

[Elektra57]

1 minute ago, leadeater said:

You are appearing in the LTT team stats

 

"Elektra57","74962988","2818","223518"

 

And you are also in the LTT Folding Month stats as well.

 

Where are you looking and what are you not seeing?

Thank you! Where did you find that. I went to: https://stats.foldingathome.org/team/223518

[Kilrah]

This only shows the first ~10K members, the team has  ~115K...

 

F@H stats are pretty limited, best to use EOC stats. https://folding.extremeoverclocking.com/user_list.php?s=&t=223518

[leadeater]

18 minutes ago, Elektra57 said:

Thank you! Where did you find that. I went to: https://stats.foldingathome.org/team/223518

Ahh.. well I run the stats script for the event so I just looked in the data for the event.

 

You can look here https://folding.extremeoverclocking.com/search.php, it's the best place to look at FAH stats etc.

[leadeater]

lol FAH doesn't like me

 

13:33:55:WU00:FS00:0xa8:ERROR:254 OpenMP threads were requested. Since the non-bonded force buffer reduction is prohibitively slow with more than 64 threads, we do not allow this. Use 64 or less OpenMP threads

[cbigfoot]

2 hours ago, leadeater said:

@GorgonHave you installed the Nvidia drivers in Linux recently? Having weirdness, first only only want basic drivers and cuda for server install. When I follow the Nvidia docs, and even HPE's docs which are the same steps, the darn thing installs gnome/x11 and I get presented with a DE after reboot... wtf? Anyway cross that bridge later, nvidia-smi doesn't detect my A40 GPUs after install.

 

I swear like 2 years ago this just worked and was dead simple.

I installed pop os 2104 got it configured fairly easily once I figured out how to edit in the etc folder. (It comes with nvidia drivers preinstalled if you choose that option)

[leadeater]

5 minutes ago, cbigfoot said:

I installed pop os 2104 got it configured fairly easily once I figured out how to edit in the etc folder. (It comes with nvidia drivers preinstalled if you choose that option)

Yea it wasn't a "software" problem, just BIOS settings. Disabling the onboard GPU or extended 3GB support fixed it (no it's not actually called that but I forget exactly what it is)

 

I use Ubuntu server, installing the drivers is simple if your BIOS isn't stabbing you in the back lol

[leadeater]

Nice got both servers configured and running, one Windows and one Linux. Windows already failed on the CPU side of things RIP. Did I make it in time for Folding Month? 

 

Spoiler

 

 

 

[cbigfoot]

48 minutes ago, leadeater said:

Nice got both servers configured and running, one Windows and one Linux. Windows already failed on the CPU side of things RIP. Did I make it in time for Folding Month? 

 

  Hide contents

 

 

 

Sorry to report but you didn’t sign up.  

[Egg-Roll]

27 minutes ago, cbigfoot said:

Sorry to report but you didn’t sign up.  

Shhh don't tell them that they might turn everything off 

 

Also I think it was more of a /s comment than anything.

[Lightwreather]

Spun up 5 quad core CPU instances since Azure just could not be bothered to properly allocate any GPU instances...

Well, here's to hoping that does something.

 

btw, we're on 1005 posts on this thread now.

[dogwitch]

2 hours ago, J-from-Nucleon said:

Spun up 5 quad core CPU instances since Azure just could not be bothered to properly allocate any GPU instances...

Well, here's to hoping that does something.

 

btw, we're on 1005 posts on this thread now.

let me know how those do. might try that next itme.

[justpoet]

Welp...last update on this crap show until I have more solid info again in probably a few days.  Back to just lame duck old CPU folding.

 

Starting to think the APU's GPU gave up on life, even though it has basically never been used and just sat undervolted and underclocked after the first month it existed.  Now even sitting at the format screen for a fresh windows install for a minute will hard lock, as does any linux install/live boot as soon as it tries to enter the gui.

 

Re-seated RAM, flashed the MB BIOS, etc to no avail.  Completely borked my linux install beyond recognition with firmware and graphics driver manipulation and LTS upgrade attempts (because why not when it is dead already) while thinking it had been a linux update that did this.  Going to truck the whole system home from the holiday trip and do re-socketing and an extensive RAM test before declaring the hardware dead though.  I'm hopeful a re-socket will help since USB is doing some odd key delay/ghosting on occasion too.

 

For Reference:

Asrock B450M Pro4 - latest AGESA (so I could drop a 5000 series chip in it if I can get one and a cheap low power gpu)

AMD 2400G

16Gig Balistix 3200 16-16-16 memory

Crucial P1 nvme

 

System normally sits remotely with an old 1024x768 Dell DVI display attached in case it is needed, but generally is run headless.

[cbigfoot]

5 hours ago, Egg-Roll said:

Shhh don't tell them that they might turn everything off 

 

Also I think it was more of a /s comment than anything.

i know... mine was me trying to be funny...

[BondiBlue]

Well, my points will once again be down today. I have three GPUs that aren't folding for a variety of reasons, so I'm relying on a few CPUs + a single GTX 1660 Ti. 

[Egg-Roll]

So I checked my GPU temps after iCUE crashed (again), apparently iCUE takes MSI's AB power settings (even when left open) and throws them out the window either upon loading or crashing... I saw 110C again and was freaking out because AB showed 25% less power still...

 

Good news, only micro bubbles and a few very small ones are left in the loop. WC nice experience, but not doing it again I want to replace Corsairs thermal junk with something good but I don't want to deal with the water.

1 hour ago, cbigfoot said:

i know... mine was me trying to be funny...

I figured but never know, it's not like we can see each others body expressions and vocal tones. Which in most cases is a good thing online...

[leadeater]

6 hours ago, Egg-Roll said:

Shhh don't tell them that they might turn everything off 

Sadly I'm only running FAH as a system readiness and thermal test along with going to compare bare metal vs VM + GPU passthrough.

 

Bare metal Linux vs Windows is currently around 15.5mil PPD vs 11.5mil PPD, a lot of that is due to that Windows being dumb NUMA problem and only 64 ish threads of the 256 being utilized. Does mean an AMD EPYC 7713 is a pretty darn impressive CPU for PPD, wow.

[leadeater]

4 minutes ago, Egg-Roll said:

it's not like we can see each others body expressions

I thought I could, swear it was something like this

 

 

[Mattias Edeslatt]

On 11/19/2021 at 11:28 AM, LazyDev said:

The GTX 1080 is the final GPU install for my system. 3 GPU's and a CPU, and my 2 420 radiators still don't get too warm.

 

 

Holy moly Batman, that GPU-sag... 

[BondiBlue]

13 minutes ago, Mattias Edeslatt said:

Holy moly Batman, that GPU-sag... 

That's connected via a riser, so it's not putting strain on the motherboard. 

[Egg-Roll]

The wheels on the bus go round and round, but not on time, as that would have been nice... Nice random BS pop up MS while I was bashing your OS  (I think it's from a program I installed while trying to get LSPDFR to work right with my controller, which failed)

 

1 minute ago, BondiBlue said:

That's connected via a riser, so it's not putting strain on the motherboard. 

Give that baby a wooden peg leg to stand on or something tho, that's cruel and unusual punishment.

 

39 minutes ago, leadeater said:

Sadly I'm only running FAH as a system readiness and thermal test along with going to compare bare metal vs VM + GPU passthrough.

 

Bare metal Linux vs Windows is currently around 15.5mil PPD vs 11.5mil PPD, a lot of that is due to that Windows being dumb NUMA problem and only 64 ish threads of the 256 being utilized. Does mean an AMD EPYC 7713 is a pretty darn impressive CPU for PPD, wow.

That's fine, at least it's going towards a good cause at the end of the day.

 

Windows has become bloatware, and a half assed stitched together operating system in general. The rate Linux is getting gamming support right now I will likely go all nix for my next build, assuming GPUs don't hit 10K a pop for a 1070 spec'd unit by then... Common Ether crash the market and flood the market with GPUs

[leadeater]

1 minute ago, Egg-Roll said:

Windows has become bloatware, and a half assed stitched together operating system in general.

It's Windows Server 2019, ain't nothing like Windows 10 or 11 lol

 

Just use Windows Server as your Windows desktop OS 

[Mattias Edeslatt]

10 minutes ago, BondiBlue said:

That's connected via a riser, so it's not putting strain on the motherboard. 

Yes, but that can't be healthy for the PCB for the GPU. I have made a support for my Vega 56 from the bottom of the case to prevent it to sag.