have you gone into advanced > configure > slots > add > gpu (or removed gpu and re-added it?
#folding-and-boinc
pastel nacelle
frozen oxide
hackintosh
pastel nacelle
ah I see
frozen oxide
boot a linux live distro maybe?
off a USB stick
since it is, after all, a PC still
split cipher
all of my school work is done through MacOS unfortunately
frozen oxide
oh are you doing school work now
because linux would only run while the USB stick is plugged in if you did it that way
manjaro I believe comes with nvidia drivers on the usb live distro but I'm not sure about AMD
split cipher
my server is currently folding, if it werent for my CPU not coming with an iGPU i would be trying to fold on it
theres not really any jobs so π€·ββοΈ
frozen oxide
I've got five jobs running now (on multiple machines) and one of them I just got 10 minutes ago, minutes after finishing one π€·ββοΈ
split cipher
i boot between windows/macos enough where idk if id reach the time limits
if only i didn't sell my old mining GPUs lol
frozen oxide
what CPU/GPU do you have?
split cipher
my desktop is a 3700X + RX580, server is 2x X5650 + P106-090
i could do CPU folding for now
mystic hollow
@split cipher and won't ever have, as Apple moved everything to Metal, no OpenCL or CUDA for you
frozen oxide
in windows you could GPU fold - on an RX580 jobs would only take 3-4 hours
split cipher
i could boot to windows overnight, i usually only ever use windows for gaming
sleek lily
cpu WU's are back, noice
frozen oxide
well the RX580 would get about 400,000 PPD and your 3700x about 100,000 tops so literally 8 hours of GPU folding would exceed the other 16 hours of CPU folding
mystic hollow
but at same time not all jobs can be moved to GPU
frozen oxide
sure
mystic hollow
when the project puts out a description like this and i am fully WTF: "These simulations are designed to test our understanding the folding mechanism of alpha-helical hairpins. We are trying to study how disulfide cross-linkers and sequence variants affect the folding thermodynamics and kinetics of these proteins, to learn how we might better use molecular simulation methods to design effective protein binder scaffolds, for use as "affibody" cancer therapeutics, for example." π
oblique pecan
Has anyone been able to get a passkey in the last 12ish hours?
tough trout
took mine over 24 hours. I requested it yesterday
oblique pecan
Okay, thanks. I'll be patient
green dove
not like there are work units to do anyway π
frozen oxide
repost but also check your spam and stuff https://twitter.com/foldingathome/status/1239264870111105024?s=09
Twitter
@green dove I just got new GPU units on two of my GPU's in the last 15 minutes (and they all spent under 5 minutes idle) but all of my CPU's have been idle for about an hour
green dove
to be fair, I think it's overloaded servers as my error rather than no WUs
Exception: Failed to connect to 140.163.4.231:80: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond
frozen oxide
yeah
green dove
I'll let it keep trying Β―_(γ)_/Β―
frozen oxide
saw lots of that yesterday but it all eventually came through
the problem I think is that when a server is overloaded then the clients are trying it even more often which compounds the issue
but it eventually clears up
tough trout
yea my station throws a couple connection errors before its able to connect and get jerbs
green dove
oh, I sit in that state for a long time, retrying every 30-60 minutes
pause/fold cycle makes it try again but fails anyway
frozen oxide
really? mine is retrying every 3 minute or so it looks like
green dove
it backs off really aggressively on my side. Unchanged from defaults on a fresh windows install
frozen oxide
weird. I just installed 2 of these systems too (both test benches) and all I did was install afterburner (to lower power limit) and folding
tough trout
niiice
north forum
i am currently uploading a WU result. its slow but steadily going. i like this a lot more then 5 failed tries and one successful one at full speed
tough trout
I'd say that's definitely better
frozen oxide
mine earlier today took many tries to upload but it usually was downloading a unit in another task at the same time so it often started even while upload was happening
tough trout
i still have a gpu uplaod and that gpu is 75% done another unit already
green dove
21:40:02:WU01:FS01:Downloading 79.12MiB
21:40:08:WU01:FS01:Download 0.39%
...
21:55:12:WU01:FS01:Download 99.45%
21:55:16:WU01:FS01:Download complete
15 minutes to grab one. But started finally!
tough trout
woot woot!!
green dove
@frozen oxide what power limit do you run at normally? I seem to not lose much going all the way down to 60-65
frozen oxide
1080ti 2-slot (in a mini-ITX) - 90%
2080ti 3-slot on a test bench - 105%
1070 3-fan (in a Ghost S1 with the side panel removed) - 90%
green dove
ah, so you're not trying to underpower them
frozen oxide
not by a lot - obviously the per-watt efficiency is probably better where you're at
green dove
mostly my office is in the sun so it's already warm in here π
frozen oxide
Basically I've arbitrarily decided I want to keep my stuff at ~68-70C if it's running all day today, that's I game in that 70-76 range so I don't want to over-stress these, so I went with power limits to match
with the ghost I was as low as 75% before deciding to just yank the side panel off
green dove
it's pretty surprising just how sensitive to temps the frequency is. Turning up my case fans to 100% results in way better performance
frozen oxide
oh yeah
green dove
even in the 60-80 C range on the card. Well within "normal"
frozen oxide
every 2-3 C modern GPU's make a step up on their frequency tables
green dove
I knew I put those 3 140mm noctua fans in the front of my case for a reason!
(it filled up the space nicely was the reason)
mellow night
anybody elses client stuck "connecting" the past few hours?
green dove
yep, everybody has trouble getting new work
pastel nacelle
bro I got scared that somehow f@h had screwed over my mouse, but as it turns out I had set my pen on my drawing tablet and it was trying to take over the curser.
frozen oxide
xD
mellow night
yeah yesterday it was making a bunch of attempts to connect before i got a WU but today i cant even get far enough to even attempt the connections
frozen oxide
@mellow night haven't seen a CPU job in 3 hours on 3 machines now... but I've had GPU jobs for the past 6 hours with only about 8-10 minutes of retries between each job (the longest was 15 minutes)
so I'd say close to full utilization of all my GPU's since I woke up
green dove
what's the 'temp limit' slider in afterburner do?
it defaults to locked w/ power limit
frozen oxide
it sets the temperature at which point it starts throttling the GPU hard
green dove
where it starts throttling hard to not go over I assume?
frozen oxide
yeah
green dove
ok, cool. So just leave that in the 80s like it's normally at, but my power limit can be lower
frozen oxide
it throttles so fast that you basically observe it sitting at that temp limit exactly to the celsius in blower/reference GPU's
green dove
and it doesn't matter since my temps are 57 right now
frozen oxide
yeah I unlinked them
green dove
this is the first time I've actually tried to understand afterburner.
frozen oxide
I've done a lot of mucking about . my mini-ITX with the 1080ti needed a lot of custom undervolting to get the best temperatures, which I threw away tho when I started folding on Friday because I was crashing and failing units... was easier to just run the stock curve
mellow night
i think this time it might be a me issue. Its stuck on "connecting" to 127.0. 0.1:36330
frozen oxide
@mellow night oh... is your backend even running? check task manager
green dove
maybe the daemon died, and your frontend can't talk to it
frozen oxide
also look if it's in the tray( notification ) area in the bottom right
mellow night
ah i did windows search to open the program and pinned it on my taskbar but when i opened it the next time using that pin i had issues.
all good now
frozen oxide
the control one is just the monitoring app even tho the icon is the same
I run control on my laptop tho so I can monitor all three, even tho my laptop isn't folding
mellow night
Is there a way to have the web viewer not refresh constantly. I keep having to erase cache to fix it but it breaks soon after.
frozen oxide
I use the advanced control and don't bother with the web viewer - the web viewer relies on their website which is having issues
though the web viewer was more reliable for me in Edge than chrome on the system that I recently built
mellow night
i usually do the same but sometimes i glance at the web viewer since it lays out the info nicely.
oh okay im using chromium edge ill try it in edge edge
frozen oxide
also I don't know if the web client can do this so
mellow night
hold up can you even access non chromium edge on w10 anymore?
frozen oxide
idk my test bench was a clean windows install yesterday - I don't know if it's old or chromium edge
but it popped up in edge with no issues
mellow night
hm ill try it out thx
frozen oxide
in either case since I don't have enough monitors to go around I just set up remote access on everything
mellow night
ayeee finally got a CoV WU
i get that they dont say the virus name in the videos for monetization but why cant we say it here???
pastel nacelle
oh shoot I am running to covid WUs (1cpu 1gpu) π
stark bough
@frozen oxide web client canβt really do that
frozen oxide
figures... so this works π
stark bough
Thereβs a mobile app somewhere that lets you check but itβs 3rd party
At least management isnβt cloud based
stuck bramble
Doing folding so far is fun did 1 job and now my cpu and gpu are waiting for a job again
STOP STEALING EVERYTHING π
stuck bramble
Lol
pastel nacelle
30 minutes
got my desk and room ready for my school's plan for distance learning. this will most certainly be something new for me. i hope that the research from f@h can help the scientists get schools back to normal again soon.
tough trout
thats why we are all here my man
mellow night
so has the viewer not worked in a long time?
pastel nacelle
yeah, on the bright side, i can leave f@h running while doing virtual classes and be here to monitor temps and stuff
mellow night
ive never been able to get it to work
pastel nacelle
also yeah its broken
mellow night
π¦
pastel nacelle
maybe with the surge of new users they will be inclined to fix it with an update that contains a covid research tab as well π
stark bough
@stuck bramble
mellow night
lolol
stark bough
Work units are rarer than toilet paper right now
pastel nacelle
unfortunately it isnt covid, but any help is help π
stark bough
Yay I got covid-19
frozen oxide
oof laptop so loud when folding... also slow... this is gonna be my one and only unit on this laptop
pastel nacelle
whould you like a gaming pc with your tp sir?
stark bough
Says everyone folding
tough trout
the any catagory has covid as prioirty right now
stark bough
Inb4 an eGPU to be used for folding is more common than TP
mellow night
@frozen oxide have you started tp get cpu WUs again?
frozen oxide
nope
mystic hollow
@frozen oxide you can access web client remotely, if configured so. but doesn't list them in one place, obviously - for every client you need separate tab
north forum
although i wouldn't mind crunching at another case
frozen oxide
did a lot of cancer-related units yesterday while they were trying to get new units ready but while I haven't been checking regularly, got four GPU's currently working on various CoV projects
mellow night
oh i just add my cpu and got one pretty quickly
oh do you mean no CoV cpu or no cpu at all?
olive olive
Just installed fah, is there anything useful to type into the team number and passkey boxes? Thanks
frozen oxide
no CPU units for me on any of three machines for six hours
noble hatch
hey, just joining the LTT team whit my 5 computers.. ;D
frozen oxide
buuut they have different intel and ryzen units... I've got only ryzens
mystic hollow
@olive olive team number is 223518 for LTT group. if you plan to do work for some other team, find their team id. Passkey is for extra security, there is a link bellow to request one (but i haven't received mine yet requested few hours ago)
olive olive
Thanks for that mate @mystic hollow, So do I need to get the key or can I avoid that for now?
frozen oxide
you can skip for now but it's worth trying to request it maybe - some people waited 24 hrs to get one
mystic hollow
@olive olive it is an extra, not required to set. Hell, not even name is required if you don't mind you don't get credited for the work
olive olive
Nah tbh I couldnt care less about credit lol, I guess I will try to get it
Reckon a gtx 1070 and R7 2700 will do a whole lot in this?
frozen oxide
the gtx 1070 yep
olive olive
I just wasn't sure if the cost in power will outweigh the gains
mystic hollow
cost in power will be a lot
frozen oxide
not really worth it with the 2700x tbh - the avx2 performance isn't great so your per-watt gain on the CPU won't be much
the 1070 will do about 8x the PPD of the 2700x and uses around the same watts - like 150w on 1070, 130w on 2700x if PBO is on
noble hatch
will this do?
green dove
to give you an idea on power, My gfx card is running at 150watts while folding. So every 24 hours, I'm using 3.6kilowatt hours. I get charged something like 13cents each for those, so folding costs me 46 cents / day if I were going constant
olive olive
SO, I could have it run on the GPU but not CPU? I don't actually care about the power cost, it aint me paying it so not my problem lol, but if its like awful power to performance then I obviously wouldn't waste time on it
green dove
cpu math can be done similarly.
mystic hollow
@olive olive yes, you can disable/remove CPU folding client
@olive olive that is possible only via FAHControl (or manual config editing) tho
olive olive
So would you recommend that I remove the CPU one?
frozen oxide
yeah just edit your slots in the advance config program
timber grove
my ryzen 7 2700x is getting about 75K PPD
olive olive
Does that look right?
frozen oxide
his 1070 will get 830k PPD though @timber grove
it's 10x... for basically the same watts
if you're concerned about power cost vs benefit,
the only CPU that are even kinda worth doing some folding on are intel 14nm and ryzen 3000 but even then they'll be beat by a 7 year old GPU per-watt
any GPU in the last 8 years is solid tho
@olive olive remove the CPU slot
green dove
my 3700x is doing 115k PPD, which is still totally blah compared w/ my 2070super doing 1545k
stuck bramble
Yes indeed @stark bough
mystic hollow
@olive olive it will work when it gets a job from the servers π
green dove
@olive olive yeah, it'll pick up work and fold it when its ready. Keep the monitoring window open and you can see it do stuff, the log tab has details
the servers are overloaded, so expect pauses between work.
frozen oxide
you can set it to fold all the time or only when idle also... that's up to you
olive olive
Ahh that sounds great, Would it stop the work if I were to start playing a game? I rarely play games but I guess its something to note?
frozen oxide
if you set it that way yes
green dove
yeah you should if the game uses the gfx card
frozen oxide
otherwise you can manualy pause/unpause it with either the web control or advanced control programs
olive olive
Alright that sounds great, My gpu is basically always under 5% load so might as well use it π
green dove
note that folding uses "CUDA" processing units, which won't show up on that perf monitor screen
click into the GPU and one of the charts, change the dropdown to CUDA
olive olive
Like this?
frozen oxide
yeah
green dove
actually not sure how much overlapping silicon there is between CUDA and 3d rendering stuff?
certainly all uses the same memory
olive olive
Ahh so this looks awesome, if FAH uses Cuda does that mean it wont be very good on AMD cards?
frozen oxide
it uses openCL on AMD
green dove
AMD has a different compute unit thingy that FAH also supports to some degree as well
olive olive
Would there be much of a benefit to me adding one of my spare RX 570s into my pc?
frozen oxide
it takes a while sometimes many months for FAH to support brand new GPU's though
the RX570 would do okay but your 1070 is obviously way faster
olive olive
Yeah ofc, I was thinking I could run them both at once
Got like 4 of them sitting around as not had time to put them in builds
frozen oxide
yeah then that's up to you if you have the resources to do it, then do it
it will even work if you can cram multiple RX 570 into the same PC you just have to set up a GPU slot for each (no crossfire needed)
presuming you have enough cooling and power supply
olive olive
I probably could do that with my 650w psu, My motherboard could actually take the 1070 + 2x RX 570s but the PSU and case would be limiting
I think I can run the 570 + 1070 no problem though
frozen oxide
running both AMD and nvidia drivers at the same time is sketchy tho
they can conflict and cause one driver to disable the other
olive olive
Ahh of course. hmm, would running 2x 570s and sticking the 1070 in a drawer for now be worthwhile?
Probably wont be in terms of power usage
frozen oxide
no, because two 570's is slightly less than one GTX 1070 in output
olive olive
Ahh right, so that will be a waste of time then. I think I will just stick to running the 1070 unless I can avoid any driver hassle
Might be able to pass the 570 to a VM and then run a second FAH client in that?
This shits probably too advanced for my skill level to be honest LOL
frozen oxide
so you're running a linux virtualization host?
olive olive
Nah, Win 10 pro
frozen oxide
then nope, there's no passthrough in windows 10
frozen oxide
hyper-V yes, on windows server 2016
but not the hyper-V for windows 10 unfortunately
olive olive
I thought you could pass through a graphics card as long as it was not being used by the host system
I take it running WinServer as a desktop OS is not a smart move?
frozen oxide
either requires linux host or windows server host
I mean, smart idk,, but you'd have to have the licenses probably
olive olive
Would rather not run linux on my main pc tbh, but Id be willing to run server if it would work as a normal os mostly
frozen oxide
you could buy another motherboard or five for less than the cost of a windows server license without a volume discount
olive olive
Well... Let's just say I don't care about paying for a Windows license ;)
green dove
@olive olive I wouldn't worry about heavily optimizing. Let your 1070 pick up some work and help out π
frozen oxide
can't say much more past that on this discord then >___>
olive olive
Unless I can get one for peanuts on eBay like you can for win 10
Yeah, I thought as much
green dove
there's so many new people the system can barely keep up as-is. Not worth a ton of effort on your part for a few more older GPUs I'd think.
join, use your best, help out, and then focus on something else.
olive olive
Ppl can get a bit funny about that stuff. Other than cost ;) Would Winserver be a viable option for a desktop OS + my FAH in a VM idea?
Would my best course of action be just to have it running on my 1070 in windows 10 and leaving it be?
frozen oxide
I mean I think it's a liability thing either way it's in this discord's rules and mods will eventually notice if the discussion goes too far down that path
green dove
yeah, just let your 1070 go for a while.
olive olive
Ahh yeah, of course. well, lets not promote anything like that then :)
This might be viable?
frozen oxide
buut yeah that - unless you have another rig that could easily run the 570 and you don't mind the power cost
green dove
installing whole new operating systems seems excessive to get one or two older gpus
olive olive
Yeah true
frozen oxide
yeah also the downtime of a fresh OS install and then a fresh OS install inside the OS install >__>
olive olive
I could maybe run it on another system but would need a new psu in there
Yeah that would be a pain BUT
If I could run that as my main OS for games and whatever, I wouldn't really mind
frozen oxide
you could just be chonking through units every ~2 hours on your 1070 though instead
rather than messing with passthrough settings
olive olive
Maybe in theory, it could help. But its sounding like trying to run both cards is more effort than it's worth
If power cost was the only limit, I would probably do it as thats another cost I don't care for... yet
Not to mention that I would have to remove the 570 as soon as I had another build to put it in
frozen oxide
well so it might work I've seen other people run 2 brand GPU but I've seen more people try it and then one or the other ends up as basic display driver
because installing nvidia copies over some of the same files as installing the AMD driver and vice versa
olive olive
Right, so for the normal user, Running AMD/Nvidia at the same time is not worth the hassle, i guess fah might be able to use a card in basic display mode but I highly doubt it ?
frozen oxide
I don't believe basic display driver is enough for folding
green dove
you'd need the specialized interface to CUDA / OpenCL, which requires the real driver
not the fallback one in windows
frozen oxide
two random nvidia cards tho which are relatively similar in age so as they can use the same driver version - works great
like I ran a 970 and a 2080ti on the same machine π
but a 460 and a 1070 β
because the newest version of driver for the 460 didn't support the 1070 and the vice versa for the oldest for the 1070
olive olive
Ahh, Shame it's AMD cards I have siting around and not Nvidia then LOL
got those 570s dirt cheap so can't complain :P
frozen oxide
yeah you can do similar with many AMD cards with the same constraints
so 5700xt and 570 in same machine no probs
but maybe not like... idk... an R7 ... whatever the number is idk
olive olive
The R7 cards are all craptastic anyways ;)
frozen oxide
it's hard to search for it because r7 is also a ryzen now
olive olive
I suppose if I were to upgrade to a 5700xt or something along those lines in the next day, that could be viable. But I don't game enough to bother wit that
I believe R7 240 is a card
Probably GT 710 equivalent
frozen oxide
ah ok
unreal rapids
Anyone know why I might be getting this error? 23:41:30:WARNING:WU00:FS00:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
23:41:30:ERROR:WU00:FS00:Exception: Could not get an assignment
timber grove
you couldn't get a work unit
unreal rapids
Right but shouldn't there be some available with the Corona thing going on?
frozen oxide
as far as I can tell nobody's gotten a ryzen CPU work unit in like 6+ hours
timber grove
nope
frozen oxide
there's still GPU work units but the servers are hammered so expect 10-15 minute waits between units being grabbed
unreal rapids
Maybe it only has WUs that can be done on GPU's only right now?
I just wanted to make sure the problem wasnt on my end. Thanks for the info guys!
timber grove
yes
frozen oxide
someone posted earlier that they were converting some intel WU's to ryzen because more people were donating ryzen power
but I assume the intel WU's use AVX512 which ryzen doesn't have
stark bough
I hear if you raid intel HQ and run the work unit on your ryzen, you printer can print out toilet paper
unreal rapids
I have an i5 and cant get WUs tho
stark bough
They took our jobs
frozen oxide
maybe those ran out too - either way three machines of mine, no CPU WU's all afternoon
unreal rapids
It started for me last night
frozen oxide
but they've all been chonkin GPU units for 8 hours with only 10-15 minutes in between roughly - and I"m not doing pause/unpause or anything, just letting it do its thing
frozen oxide
I did have one job kinda-download and then not also so there's more issues for sure
unreal rapids
Maybe because the college is closed because of the weekend and Corona
olive olive
Anyways, I am going to go to bed now. Thanks a lot for your help @frozen oxide @mystic hollow, Have a good night
My college is meant to be open despite it RIP @unreal rapids
north basalt
anyone with the unraid docker know how to fix the time variable?
my logs all show in UTC
green dove
switch yourself to UTC
green dove
not even kidding, my programming terminal window has local & utc time. Easy-peasy
north basalt
in unraid, the correct time is set. just not in the container
green dove
TZINFO=America/Denver or soemthing may do it as a passed in env var if you really care
the time is set, but just off in utc timezone instead right
north basalt
right
in unraid i have it correctly set to America/Chicago
in the container its UTC
green dove
try TZ=Europe/Amsterdam except chicago
as a new env var
assuming the app even logs in local time vs utc time.
green dove
Just TZ it looks like. but it's really easy for app devs to force users into utc for ease of correlating logs later
ie, you complain about no uploads, they can match up times on the server's utc based logs easily
so who knows if folding actually listens to that variable.
green dove
don't know then. Does it matter?
north basalt
just for troubleshooting and/or seeing when i can actually get WU versus something else wrong
green dove
you'll be fine. Some googling shows forum threads from like 7 years ago saying: "yep, logs are in UTC" Β―_(γ)_/Β―
green dove
meanwhile my upload is failing for like 20+ minutes.
north basalt
im guessing no one is @ stanford to fix it
green dove
starts, gets one upload tick in, fails.
nah, it's overwhelmed servers, not much a person at a keyboard can do
north basalt
maybe not at a keyboard, but a person in a datacenter could.
green dove
vs. spinning up new ones.
perhaps, maybe. Depends a lot on their architecture and if they even have physical servers ready to go
north basalt
true
frozen oxide
I just got more WU's on 2 of my GPU's sooo they're out there frens
snow escarp
My GPU WUs also picked up, servers seem to be better now :D
green dove
@frozen oxide what IPs? I can't get an upload to stick for an hour or more.
much less download new work units
Finished: 23:51, upload finally succeeds after many attempts: 00:42.
I wonder if I just am on an iffy server. No clue how they shard users
frozen oxide
can you override them anyway?
155.247.166.220:8080 is a work server I just got a unit from 20m ago
155.247.164.214:8080 25m ago
128.252.203.10:80 10m ago
green dove
00:20:01:WU02:FS01:Assigned to work server 140.163.4.231
can you? I haven't looked for a setting
frozen oxide
idk I've never done anything weird because the longest any system's been without a GPU WU for me is 20 min
green dove
cool, I'll go look after dinner. That server in particular seems to be down or harder hit
frozen oxide
yeah I mean the log shows a bunch of those connection errors
but with no intervention from me it's eventually finding some work
in many cases the new WU is downloaded as the other one finishes uploading
timber grove
oof
frozen oxide
you should tweet that at fah's tweet that they're getting blocked for spam lol
upper shuttle
aight
i have no idea how its sending me an email at 1am when i requested it like almost a day ago
Twitter
only one comment in that thread has the same issue as mine
upper shuttle
omg, its still sending me https://img.thaun.dev/ckwlb.png
north basalt
show off
upper shuttle
oh and wtf is this "next attempt 35 minutes"
north basalt
my gaming PC is folding but my unraid backup server isnt. lame.
upper shuttle
my ryzen 7 pro 3700u cpu is like on idle :/
though my other intel i5 cpu got a wu instantly
north basalt
when i restart it, it gets a WU. then does nothing
lucid plinth
rich depot
what's Project 14198 for?
jolly jacinth
where is the team
tough trout
Team number is - 223518
hasty mirage
give it some time
jolly jacinth
aight
hasty mirage
took a couple of minutes when i signed up
jolly jacinth
I'll probably spool a vm up on esxi too
well no
2600x wouldn't do too well would it
doesnt have a gpu
tough trout
it would do okay, there are cpu specific tasks
jolly jacinth
covid?
tough trout
all in the FAQ my friend
jolly jacinth
Just did some reading and while im waiting im folding anonymous
my cpu is sitting at 29% and gpu at 4, opened advanced and tried to figure it out, no luck
tough trout
did you actually get a job?
tough trout
looks like its working
jolly jacinth
my 3900x is doing 13833 and my 2070 is doing covid19
11748 specifically
so im just waiting for my passkey?
tough trout
are you on windows?
jolly jacinth
Yes
hasty mirage
still haven't recieved it?
jolly jacinth
Nope
hasty mirage
hmm.. odd
tough trout
it is taking a long time, servers are stressed with all the requests
hasty mirage
probably
sacred spruce
has anyones app crashed? i downloaded it today and completed one round good but now i have nothing downloding
hasty mirage
or check and see if it accidently ended up in spam/trash folder or something
tough trout
in windows you can right click the folding icon in your system tray and go to advanced control and input your key there when you get it
hasty mirage
check your logs and see if you get errors connecting or no jobs available for your system
tough trout
@sacred spruce it is taking awhile for jobs to be handed out today just leave it running and it will fix itself. Or try restarting the app once. Only once though
sacred spruce
check your logs and see if you get errors connecting or no jobs available for your system
@hasty mirage yea no jobs are found
tough trout
There is plenty to get around, just have to be patient
sage bay
No patience. Must go faster.
hasty mirage
anyone else getting bad gateway when accessing the stats page?
sage bay
Nope
sacred spruce
anyone else getting bad gateway when accessing the stats page?
@hasty mirage yes
tough trout
yea its been down most of the day
tough trout
that is another 3rd party stats page that is slower on the updating but is more reliable
hasty mirage
awesome. thanks for the link.
sacred spruce
uhh
@jolly jacinth use another browser i use edge and chrome
sacred spruce
I've been having to use edge
@pastel nacelle same
pastel nacelle
it seems they are doing better at rolling them out recently
pastel nacelle
yo lets go! im running a gpu and a cpu covid wu!
sage bay
nice
hasty mirage
cool
tough trout
Good work peeps
sacred spruce
i completed one job in 2 hours, instead of 5 with an RTX 2060 upgrade i had 3g 1060 last year when i started folding
sage bay
I'm waiting for my shiny new 5600xt to crunch some numbers π
pastel nacelle
sacred spruce
changed my disease to cancer to see if i get any numbers
tough trout
probably wouldn't make a difference
sacred spruce
oh well yea i cant get any assigments till tomorrow after work
frozen oxide
basically all the WU's for other causes were completed like friday into saturday - I don't think new ones were added - they were quickly adding more CV units though
fierce steppe
Been stuck waiting for download for 7 hours now. RIP.
timber grove
oof
thick holly
Mine is still working somehow.
green dove
I have not looked at boinc yet - does that have CV jobs? Or other stuff worthwhile?
noble hatch
Been stuck waiting for download for 7 hours now. RIP.
@fierce steppe i did a reboot then it worked.
LTT is raning #89 wold wide team in total Bonic credits
in Rosetta and Covid-19 folding, the LLT goup is polling only 200k a day. the members should conentrate more on the Rosetta Project.
mystic hollow
@jolly jacinth my 3900X have yet to get a second job
scenic dragon
"ERROR:Receive error: 10053: An established connection was aborted by the software in your host machine." does this mean my firewall blocking the client?
supple laurel
anyone know the covid-19 work numbers?
fierce steppe
Covid-19 work units are 1174X
Project 11752
SARS-CoV-2 (COVID-19 causing virus) receptor binding domain in complex with human receptor ACE2 (alternative structure to 11741)
Also looks like they might have added some new ones
1175X projects seem to be new Covid-19 units
unreal rapids
Anyone know why I am getting this: Exception: 10002: Received short response, expected 512 bytes, got 0
unreal rapids
Hmm okay. I am getting that error on my GPU
north forum
everyone gets it from time to time
unreal rapids
I am getting this for my CPU: Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
north forum
jsut means that the F@H server is emtpy and/or has a hickup
same here
i am getting gpu units quite frequently
but no CPU
nothing to worry about
unreal rapids
Okay thanks for the help dude
north forum
npnp
sleek lily
damn. cpu WU's were often running dry the last few days
But now even the GPU WU's are out
... guess that's a good sign xd
rich turtle
mystic hollow
decided to start my 3700x system too, but only for gpu tasks (CPU tasks will be handled by my server with 3900X)
no jobs for anything anyway π
sleek lily
@jolly ether if you click on the gpu, you'll probably see that it doesn't show any project. That's probably due to the shortage of work units right now
my cpu and gpu both don't have any work units to work on either
we'll just have to wait for Standford to generate more of them
upper shuttle
they are still sending me emails :/
jolly ether
Ohh ok just making sure I wasn't doing anything wrong
north basalt
plenty of emails and no WUs here. im not building a dedicated folding box until i constantly get WUs
upper shuttle
i dont have the money to build a folding box
jolly jacinth
still havent gotten my passkey
mystic hollow
i have two CPUs (3700X and 3900X) plus a GTX 1060 6GB, plus at first test i had a 2950x and 2080 Ti also running and in a day i got a total of 5 jobs - one GPU job, 4 CPU jobs.
and yes, passkeys are probably arriving in a week or two π
jolly jacinth
jfc
rich depot
This is my folding box, with a 5960X
frozen oxide
that'll probably go to 900k with a passkey
jolly jacinth
yeah, which I've been waiting all night for
mystic hollow
big piece of nothing, no passkey
frozen oxide
there's no work to do anyway
timber grove
i have two projects that are trying to get sent out for 9 hours.
frozen oxide
same I have a 5 hour failed send
maybe they ran out of disk like they were talking about on the forum
no room to store all the uploaded results
timber grove
they need to ask linus for a storage lone
frozen oxide
yeah but then he'd be too busy to show up for a year and do the video so the HDD's and storinator would be sitting for 8 months
@mystic hollow I'm surprised tho you go so few jobs yesterday, I got 20 WU's done in 9 hours yesterday (with 3 systems of similar spec) all but two WU's were GPU
rich depot
I imagine their bigger issue isnt storage but infact it is bandwidth and sheer processing power
frozen oxide
@rich depot nah the sysadmin literally said on the folding forum they were running out of storage space
and they're working on moving stuff to the cloud but I guess uploading stuff from their lab to the cloud is a new bottleneck
rich depot
Can you link me to that please? I'm trying to keep up with it all
frozen oxide
rich depot
Thank you! I'm usually on top of this stuff quite well but getting ready for the event is taking it's toll
I'm actually friendly with someone over there so I'm going to reach out and see whats what. But it seems that they havent run out of space yet, but are expecting too. This problem is they just dont have the computation needed to generate the WUs quick enough
timber grove
what is the most reliable way of seeing how many work units you have done
frozen oxide
the stats page, which is not working right now
timber grove
whats the next best thing
north basalt
bad gateway]
timber grove
other than ECO
north basalt
my team wasnt found there
frozen oxide
only top 6000 teams on that database
frozen oxide
my other team hit top 6000 yesterday morning so the history on EOC is only since then
dusky loom
@frozen oxide why another team π’
north basalt
boo im a new small team
dusky loom
It kills me: I am getting GPU jobs on my 48 thread monster on my K5000
north basalt
im getting no GPU or CPU jobs rn
frozen oxide
you can easily see where new WU's were added saturday night just like they show in the forum and then dropped off again
north basalt
on 3 diff xeons (V3) and 4 GPUs
frozen oxide
nobody is
dusky loom
#goals
frozen oxide
ryzen CPU jobs ran out about 24 hours ago, intel CPU jobs about 18 hours ago, and GPU jobs seemingly about 8-10 hours ago
north basalt
good problem to have i guess
timber grove
i restarted my pc and got 3 jobs within 5 min of restarting
dusky loom
The gpu job I have is very old, but still going
mystic hollow
@frozen oxide to be fair, that one WU was nice
sleek lily
Be so cool if Linus could do an emergency petabyte project for Stanford (F@H).
"Youtuber helps Standford uni with server problems" XD
stark bough
that would require traveling there for deployment. And dude just lost his cat
Might be a good video for networking and distributed computing though
Iirc itβs all centralized anyway, to maintain intregrity of the data
People tend to take edge servers for granted
sleek lily
"that would require traveling there for deployment."
I know, but on the other hand, wouldn't be surprised if he could get some sponsors going for a scientific adspot sorta
and yeah, had already read about his cat
rich depot
I have multiple RTX graphics cards that I am currently not using, I also have an extra i9 that I can piece together for a build, should I assemble it to run folding@home? Like does it actually help?
I would be more than happy to run it at the cost of electricity. As long as it helps
my extra RTX cards are as following;
RTX 2060 FE
RTX 2070 Super Jetstream LE
RTX 2080 Super FE
RTX 2080 Super GameRock
RTX 2080 Ti Ventus OC
I have the Aorus Master Z390 Mobo, so I think I can use two or three of those cards
stark bough
@rich depot Every bit helps. At the very least it adds heating that would have gone through a heater anyway
rich depot
can it run on different graphics cards?
like can i assign a worker on the 2080 ti and another on my 2080 FE?
stark bough
Way I see it.
I could put 700 W through a heater, or through my system and do actual work
stark bough
Yes, CPU/GPU/ all get their own work units
mystic hollow
except Mac. No GPU support on Mac
stark bough
Theyβre paralleled
rich depot
K, which GPUs should I throw inside my PC?
stark bough
work independently. No SLI here
rich depot
yeah, not expecting that either
stark bough
Whatever you feel provides the best efficiency imho
rich depot
I honestly don't care about 30-40 euro of electricity.
stark bough
We got so many rigs thrown at the problem that work is actually hard to come by, and the points donβt matter so do the most efficient job
rich depot
lol really
stark bough
In general higher end processes will help. Iβd honestly do boinc as there are other. In covid projects that can use the work
I just run Folding cause I donβt have high end hardware anyway
stark bough
Sure. Iirc x4 PCIe 3.0 minimum.
No risers here
There are also videos on why folding is such an important problem to solve if you donβt mind some links
outer solar
Quick question, does anyone know if changing my identity within FAHControl app will alter my points and stats?
mystic hollow
you will be a second contributor in eyes of stats
i mean you will be two separate users
frozen oxide
shall i? fold?
@small current if the laptop has a discrete GPU (like nvidia/AMD) yes. If not, then don't bother - even a 16-core CPU can't beat a GTX 1050 and there's no CPU jobs available the last day - but the GTX 1060 max-Q in my laptop can nearly double the work output of a 3950x ryzen so it's still worthwhile.
small current
Yes it has a 1060 max q
frozen oxide
Felaneous - then yeah for sure - I am using a 50 mhz underclock on my 1060 max Q to keep it under 65C - work unit took 6 hours and the laptop was loud, but it didn't keep warming up so I was fine with it. Also disable the CPU slot so you get the most thermal headroom for the GPU
also keep in mind the power brick gets very warm on the gaming laptop when folding so don't have it sitting in the sun or trapped under stuff
mystic hollow
@frozen oxide not like there are GPU jobs either
frozen oxide
got a new GPU job less than an hour ago π€·
frozen oxide
I'm working rn tho so only the one test bench is running - their post 12 hours ago seems to suggest the WU's are out there but their servers are hammered trying to hand them out
small current
I am currently gaming so I won't fold now
frozen oxide
idk feel the power brick when you're folding after about 30 minutes and if it's like ouch-ouch hot then move it somewhere with either airflow or cooling - but if you can game on it and it doesn't overheat it's probably fine when GPU folding only
just note that you only get credit for finishing a whole unit and my 1060 max-Q took about 6 hours per unit. if you pause it your credits per unit will go down until you finish it (incentive to finish jobs)
so like if you took 3 hours of work then paused it then finished it tomorrow you'll get less points than if you did it 6 hours straight
mystic hollow
@frozen oxide it doesn't go down afaik if you hand in the result before expiration. bonus points might, but for that one would need passkey, which most people can't get at the moment π
frozen oxide
yeah bonus points sorry
it's a shame they are so wonky with passkeys, like some people getting 40+ emails and most people not even getting one
rich depot
@stark bough Do you have any idea how I can get a CPU task?
north forum
Just wait
rich depot
I recovered my mining rig with an i3 8100 and 8x GTX 1060. 6 of my 1060 got a task, but my CPU still can't find one
rich depot
oh alright, do I need to tune anything in order for it to get one faster?
all my 1060s got a task.
frozen oxide
CPU tasks are scarce
stark bough
just wait for it. WU's are more readily avaliable than toilet paper
north forum
no, you can't force it
frozen oxide
ryzen CPU jobs ran out about 24 hours ago, intel CPU jobs about 18 hours ago, and GPU jobs seemingly about 8-10 hours ago
rich depot
Lol, I ordered like 8x 10 rolls of toilet paper, it got delivered last Friday
I just didn't want to go without toilet paper
stark bough
where are you keeping track of these
frozen oxide
the GPU jobs are back again since I said that tho... but the CPU jobs nobody's posted about making more
north forum
i got one cpu job today
frozen oxide
I'm just watching what people say in a few discords plus the forums
rich depot
should I set all my 1060s to max fan speed?
I have a noise insulating room where I keep my servers, noise isn't an issue
frozen oxide
@rich depot that's up to you... but if you can stand the noise more cooling can't hurt
rich depot
but does it damage the bearings
north forum
if you don't care about noise, thenyeah, why not. the cooler the gpu's are, the more freely they can boost
rich depot
it's the FE 1060s
frozen oxide
I doubt it's any damage to run the fan at its rated speed
rich depot
got it, to max we go
north forum
running them at 1005 is within the expected limits, so it should never shorten the lifespan of the components
100%
frozen oxide
I mean if we're talking you're gonna run this months from now, you should do some of the things miners do and lubricate your fan hubs or keep fans in spare
rich depot
it has served me well mining bitcoin, I see that it will just fold from here on forwards haha
mystic hollow
to be fair, typical folding load is around a better game load
rich depot
lubricate fan hubs?
rich depot
95% currently
frozen oxide
idk I'm not a miner but I've seen posts of people servicing fan hubs - no idea if it actually helps or not
mystic hollow
so it is nowhere near the load mining put on the card
rich depot
66 degrees
frozen oxide
idk temperature is the best indication of the load
rich depot
with 80% fan speed
frozen oxide
is it warmer or hotter than when you mined on them?
frozen oxide
then it's probably fine lol
frozen oxide
also underclocking is always an option - I am running -50mhz and 90 power limit on my mini-ITX - that tiny chassis cannot take hours and hours of load
rich depot
yeah I can't be bothered underclocing
because I need to do it to all 8 of those 1060s
frozen oxide
not the tiny curve settings
rich depot
and frankly, I can't be fucked
rich depot
will it crash
frozen oxide
not undervolting one frequency at a time
frozen oxide
nooo
rich depot
hm
frozen oxide
nah undervolting is too tedious plus my undervolted 1080ti crashed on the first unit it did
rich depot
got it, all -50
frozen oxide
but like... I got 6 degrees less and only dropped my mini-ITX from 1800 to 1750 mhz - granted that's a cooling limited chassis
rich depot
should I keep the fans running at 80% or should I let it go down to 30% then ramp back up when there is a WU?
I heard ramping the fans breaks them
frozen oxide
idk I'm no expert on this and I don't intend on folding forever - just a few weeks
rich depot
got it, I'll keep it on 80% since I have minimal dust in the server room
stark bough
really depends on the fan type. rifle bearings require more lubrication iirc. dunno of fluid dynamic or magnetic really requires it
rich depot
it's the blower style fans
Nvidia FE GTX 1060
I got like 8 of those for mining
they've been mining non stop for 3.5 years now lol
stark bough
got it, I'll keep it on 80% since I have minimal dust in the server room
@rich depot
I just force filtered air into my intakes
rich depot
I filter the air in my entire server room with a 2000 buck air filter lmao
mystic hollow
@stark bough what did you do to that poor case
weary epoch
anyone else getting 'No WUs available for this configuration'?
weary epoch
ah well, I suppose they'll fix it soon
rich depot
all my GPUs are working though
weary epoch
CPU and gtx 980 ti can't find a task
rich depot
I have 8x 1060, all working
weary epoch
I have one really old first-gen gcn firepro w2100 working, just because it takes several days to finish each work unit
rich depot
lol
weary epoch
everything else gets 'No WUs available for this configuration'
weary epoch
yeah, but I'm just kinda sad the bulk of my compute power (gtx 980 ti, gtx 1080) have no WUs
stark bough
thats weird. Copy is at 48%
weary epoch
you know what, right as I said that, the 980 Ti got a task.
I guess there's more GPUs out there folding than there are work units
rich depot
yeah, there has been a tremendous amount of support I believe
stark bough
@mystic hollow im literally using it as a forced air room heater for my living room so gotta have that airflow
they need to push people off to boinc so every other project can get their WU's cleared out
rich depot
Btw guys, I have the corsair 1600W power supply to power an 8100 and 8x 1060, I'm using a bunch of splitters to get enough PCIE cables, is that safe?
frozen oxide
@weary epoch GPU WU's are out there but they made a post about 12 hours ago late last night saying that their servers are hammered trying to give them out, please be patient. As far as I can tell though no new CPU WU's have been added and people are only getting a handful of "leftover" ones that are probably expired ones that other people didn't finish in the last few days
stark bough
i wonder how much idling compute capacity is on F@H just being bottlenecked by the work that actually is avaliable
sounds like distributed computing needs some more distributed serving
weary epoch
@frozen oxide thanks, I didn't see that post. Do you have a link?
rich depot
Isn't it that you are less likely to get a CPU WU if it's set to a prime number amount of cores?
such as 5,7,11
weary epoch
Not sure, but I have it set to an even number of threads so idk
rich depot
it's recommended I believe, to change the threads or cores to 4,6,8,12,16
frozen oxide
when mine was set to 5 (since default is N-1 ) on my 6-core... it showed in the log that it was changing the N count because it was prime
and lowered it to 4
so I don't think it matters what you set it to since the worker will adjust
rich depot
I'm not offering my server CPU lol
frozen oxide
sorry it was 11 and it changed it
rich depot
@dusky loom try 48, I don't think 47 is good enough
dusky loom
I am π it was just sitting in my garage off
46 or 47 (i heard having gpu requires a core)
frozen oxide
according to forum
The default settings after installation of the client run it at the Medium setting on a desktop (Light on a laptop). For your system that would be one less core or thread than the total available, or 23. Because that is a large prime, the folding client will back that off to the largest number that is not an odd number or the product of a prime number 11 or greater. So 22 and 21 will be out and the client will use 20.
dusky loom
oh this is automated, so I don't have to deal with it
frozen oxide
yes
dusky loom
bumped it from -1 to 48
frozen oxide
well, to be fair, people on 4-core i5's with 2 GPU's installed and no CPU slot at all were seeing almost 50% utilization because there is nearly a whole core used feeding the GPU with data @rich depot
stark bough
the real problems with core count and GPU's is multi gpu systems with not enough cores. i.e 4 GPUs, on a 2 core system
as most people don't typically run multi gpu, its 99% not an issue
frozen oxide
these were slower i5's and I don't know if anyone's checked on like a 9900k or 3000 series ryzen how much of the core is being used
actually I'll check now since my CPU is basically idle
rich depot
strange
How can I let one of the GPU be the iGPU while the other be an MX 150?
this is for my old laptop
@frozen oxide ideas?
dusky loom
ugh, I'm not getting jobs and losing rank! π’
north forum
@frozen oxide i have a 9900k, my last CPU WU today took around 60% of my CPU power overall with 14 of 16 threads enabled for F@H. but i also have seen CPU WU's that literally eat ALL of my CPU power
frozen oxide
I meant how much of a core is being used by the GPU worker
dusky loom
I was hoping to hit rank 6000s
frozen oxide
looks like about an entire CPU thread on a 3900x is being used by fahcore_21 - works out to 3.9% but on the graph it is actually split on half utilization of two threads - so on even slower CPU's I can see it pegging a whole core
stark bough
we be headed to the border. rumor has it they got work down there for people like us.
frozen oxide
so 2-ish threads at half utilization are feeding my GPU with no CPU slot enabled at all
rich depot
how can I make my IGPU work also?
dusky loom
As these aren't MPI style jobs, I am going to step it up to 48/48 cores and the gpu will get distributed out
I can't imagine the poor k5000 taxing the CPUs too hard
frozen oxide
afaik they don't have a driver for intel igpu @rich depot - it needs to support openCL or CUDA
rich depot
ah
frozen oxide
maybe it works on the newer ice lake igpus
frozen oxide
those new intel GPU's are being sold to datacenter for compute so presumedley they have the right compute API drivers
the same is probably true of ice lake based on the demo from adobe at CES maybe
yeah so probably only 10th gen and newer
stark bough
@dusky loom https://www.youtube.com/watch?v=jz_GckHujqs
People trying to get rank during CV folding hype
dusky loom
I'm bustin' up over here
dusky loom
The F@H algorithm that saw my K5000 passing over other's gpus... I don't get it.
I'll get it done in thrice the time at 4 times the power cost.
SOLD
stark bough
"look mah, im helpin"
rich depot
my laptop isn't getting tasks
dusky loom
@rich depot I've got a lot of hardware just sitting
rich depot
Imagine if they would give empty hardware stuff to mine and collect the money for that, then place it on research
cloud mining
my CPU found a task!!!!
it likes my 8 core laptop than my 4 core mining rig lol
warm plaza
I guess BOINC's Rosetta project is also COVID research
According to https://unraid.net/blog/help-take-the-fight-to-covid-19-with-boinc-or-folding-home anyways
stark bough
im wondering about egpu configs
mystic hollow
@rich depot you ask for work for iGPU... i ask for work for GTX 1060 GPU and 8-12 core CPU. But F@H clearly hit recession, so no work for anyone π
stark bough
just ran out of work for both
mystic hollow
warm plaza
Ii'm chugging through a 14328 CPU unit
north basalt
at least ur workin
stark bough
are you seriously running Distributed on your nas π
frozen oxide
@mystic hollow did you try pausing and starting the specific slot in the advanced client?
dusky loom
NAS has 16 threads, idles using like 1
mystic hollow
@frozen oxide everything, even restarts. There is nothing
stark bough
2016: People running GPU's to find bitcoin
2020: people trying to find work for their GPUS
frozen oxide
I cant stand the noise in my office so I won't be running any more during the workday - hopefully there's some overnight units for me
the fans aren't the issue it's actually the coil whine lol
went and stopped all the fans on the test bench and it was actually the whine
dusky loom
I find the MLs in my Desktop cover the whir of the server downstairs nicely when both are loaded
You could hit the chokes with some hot clue
frozen oxide
nearly any paste will do fine - in a chart of 70 pastes tested, the best paste to the worst paste - they're all within 3C
even denture cream is within 5C of the pastes
liquid metals are several celsius better but not really recommended for long term use
timber grove
use some tooth past
stark bough
tooth paste if for broushing the CV off my teeth π
rich depot
I just got a massive CPU WU
it says it'll take my 9900k 10 hours to complete
and that time is going up
warm plaza
Which project?
rich depot
14051
Dysregulation of Src-family kinase (SFK) activity is observed in many forms of cancer, contributing to tumor cell growth. SFKs are involved in a variety of signaling pathways that are found in virtually every cell type, including innate immune response, cell proliferation and survival.
whatever that is supposed to mean
stark bough
if base = estimated, its probably still calculating.
warm plaza
I should look at bringing my storage array into the business but I have to use Linux compatibility layer on FreeBSD and I'm lazy
stark bough
@rich depot Kinease is involved in metabolism, aka how cells run. ATP is a unit of metabolism in cells, like Watts in a PSU.
Researching how that 'power' is distributed, and finding the paths goes a long way towards fucking with cancerous cell growth, kinda like snipping the right wire on a bomb, or knowing that a disconnected RGB cable isnt necessarly going to cause a system to shut down
warm plaza
I wish this was a BOINC effort instead of FAH, every time I use FAH it seems so.... chintzy and terribhle
stark bough
its relatively 'easy' to read DNA strands, but hard to predict how a 'printed' strand folds into a 3d shape, cause the '3d' shape dictates how the protein functions.
rich depot
ah
stark bough
you probably seen things like these. GATC repeted over and over. Thats bascially the source code of everything. DNA.
Folding at home is basically cutting out specific strands, then 'compling them' to see how they work
since we can rapidly sequence things, but necessarly know what the 'compiled code' looks and behaves like
Central Dogma of biology.
linear 'printed' code is spit out, and then the whole damn things folds in on itself.
ive been needing a refresher on how Folding @ home relates to the principle of how it all works anyway
rich depot
interesting, it can do all that with a CPU+
?
I haven't worked with biology since like forever, I work in AI, and those look as much alien as they do pictures for me
stark bough
Sprare CPU cycles
Iβm more familiar with the biology than how things are chopped up for processing. But yeah itβs basically prediction of interactions
You start out with these stores in the nucleus
They get copied, sent off to a ribosome, and 3 base pairs translate to an amino acid
rich depot
It is basically modelling on an atomic scale. We know that various aspects of the molecule are hydrophobic, hydrophilic, like to bond to certain other atoms ect. FAH takes this information and runs simulations changing attributes such as acidity, osmotic pressure, 'miscoded dna' ect and then projects how that protein responds. They also simulate how different proteins interact, this is the kind of simulation we are doing for COVID. A wu is a tiny fraction of this whole simulation. It is done on the fractions of a nanosecond kinda level.
stark bough
You can kind of see how itβs like programming
A lot of those terms are hard to understand for someone not versed in chemistry. Trying to explain it through how it all ends up there
But yeah, proteins fold incredibly stupid fast, a WU is basically covering a few nanoseconds of an incredibly large and complex molecule interacting with itself
Inb4 a different level of abstraction through enzymes
rich depot
WUs are actually often fractions of a ns which is crazy when you think about it
I think there is just so simple way to explain the simulation other than saying it simulates the way atoms move. Anything more in-depth than that is too technical for most. I like to think I have a decent understanding of how it works in theory, but I wouldnt say I truly understand it if that makes sense
stark bough
@rich depot Either way
-Biology is incredibly complex
-The people who figure this stuff out are basically geniuses, the ones who program work units even more so
-The overall premise of the work we do is totally worth it
rich depot
No simple way*
And yes I agree, people who can not only understand this, but also somehow translate it into computer coding are truly gifted people.
stark bough
its hard enough to get people to understand enzymatic actvity, let alone how variables in ph/cncentrations/tems work
it takes weeks or months for people to wrap their head around translation alone
and in the end it all takes nanoseconds to do
https://i.pinimg.com/originals/fc/92/57/fc9257c7df044688f34514687085e827.gif
cause it all basically shoots out like this, folding in on itself as it comes out
rich depot
I have spent a few years reading up about it thanks to F@H and I can barely grasp it honestly
rich depot
F@H has surpassed 6m WU requests per day. over 70 per second. Ad it's still going up
plucky smelt
wow! no wonder I can't get any WUs
rich depot
They are currently rolling out more sever capacity. They have the WUs now but don't have the capacity to distribute them
plucky smelt
Good to know
north forum
dusky loom
Anyone know off the top of their heads why there is relatively low ram usage
north basalt
when folding?
dusky loom
yeah
I have messed around a little bit with HPC and most of those nodes have ungodly amounts of ram
rich depot
Because we are doing TINY fractions of a full simulation
stark bough
to do a full molecule for 200 ns would probably blow up your computer
rich depot
Ha, it just wouldnt run. Imagine the page file on that thing!
stark bough
Help theres amino acids pouring out of my SATA ports
rich depot
-.- look at the pinned posts
restive lance
tfw did but wasnt obvious
north basalt
im getting WUs but now for other stuff
rich depot
Click on the first link in the pinned post, all the info you need is there @restive lance
restive lance
ye i know i found it but the first time i did it i couldnt see it
rich depot
No worries
warm plaza
team0 has an incredible amount of points/24h though
mystic hollow
@stark bough literally got one CPU and one GPU job all day https://folding.extremeoverclocking.com/user_summary.php?s=&u=873758
warm plaza
my one machine is idle, the other is working on a longer cancer one I think
both are CPU-only
stark bough
RIP
@mystic hollow you have no future
inb4 people VPN to silicon valley thinking it gives them priority in folding
supple laurel
hey boys turn it up we only need 350million credits to pass team 0!
stark bough
Didnβt we already beat team 0? You talking week?
supple laurel
month
stark bough
Also the bottleneck is their ability to distribute WUa
tall steppe
Hey! Just joined wondering if anyone knew what was going on, I'm not getting any work units. I saw a forum posts where others said the same, so it's not isolated. My setup is old, but just thought I'd jump in and take part in this.
supple laurel
there has been such a outpooring of support the server can't keep up in sending out work units.
tall steppe
Ah yes, I saw that mentioned also. Hey, it sounded like CPU WU was separate from GPU WU, is that true? As I said, my desktop is old but maybe I can switch it to CPU mode at night at least if it's possible to get WU that way.
supple laurel
they are separate. you can run them at the same time if you like. depending on your setup. the last 24hours or so i've only been getting GPU WU. but before that it was only CPU WU so it's kind of 5050 what you can get right now
mystic hollow
@stark bough btw passkeys: "Once Google decides there are too many passkeys being delivered, their anti-spam software kicks in and they start blocking our emails. Then it takes person-to-person contacts to get permission so send an email to your gmail account."
stark bough
must have C.Virus to research C.virus
So i just sneeze on my GPu now?
@mystic hollow
supple laurel
i need to get some WU my room is getting cold....
supple laurel
11C.....
stark bough
you poor dear. you need work
supple laurel
the plus side my idle temp is down to like 35C for a 4.2 at 1.4v 2700x
lilac matrix
sharp gyro
I have not received a Code that gets emailed... Are they having issues getting those out as well?
supple laurel
are you using gmail?
sharp gyro
charter
supple laurel
there was a link somewhere up top about if they get pinged for to many code emails google starts marking them all as spam
sharp gyro
yeah, I checked my spam folders.
I had one sent to my gmail as well when I didn't receive the one for my charter account.
supple laurel
they might be trying to say that google blocks the outgoing email as a spam account.
sharp gyro
gotchya, It seems like they have had a few issues from the influx. Will give them time.
glacial mica
8 hours without a a WU asigned. Time to sleep to me I think I'm going to give the rig a nice sleeping time.
oblique pecan
I have WUs on 2/3 of my GPUs
sharp gyro
I don't use my PC that often, So i just said screw it and decided to fold on my main and and card I had laying in a box
Might make it useful and move the PC to my bedroom where it is 10 degrees colder than the rest of the house
timber grove
does anyone know why my computer can not send projects 11757 and 11756. my computer has been trying to send them out for over 14 hours now.
hearty rampart
Threw a $50 steam card into the giveaway prize pool. Fold on, guys. Youβre doing great work.
green dove
Is there a site for seeing my own stats that works? The official site is not loading
rich depot
@silver wing Ignore this
timber grove
ummm i have a cpu project on a gpu. the project is 11761
fierce steppe
I believe that is an error in the description
- How do I know that I'm working on COVID-19 related projects ?
If you're running a project that's listed here, you're fighting COVID-19 :
CPU : 14328 - 14329 - 14530 - 14531
GPU : 11741 to 11764
rich depot
Finally got something
bitter cedar
what can cause the client to not start back up after you quite it
limpid pilot
Once you pause, after a couple of minutes the config is saved. So maybe you paused, config saved, and you relaunched it where it maintained the same state?
(you can check the log tab in FAHControl to see when it writes out the config file)
thick holly
Since I am not getting any more jobs, should I quit the client and start it back?
tall steppe
Hasn't helped me. π
stark bough
@hearty rampart did someone seriously start a folding pool?
cause ltt needs to beat PCMR in GPU units
hearty rampart
IIRC we're #2
stark bough
someone should really pin our stats
hearty rampart
In users that is