#folding-and-boinc
1 messages Β· Page 52 of 1
ah yes i got it thanks
There is a trick to find a WU before the current one finish.
restart client
Just found it
how do i restart just close in task then reopen?
so for some reason despite completing 2 WUs yesterday, my client still says the last completed WU is from 5 days ago, and it hasn't added my points from yesterday either
if you're talking about the web client its buggy
the web client hasnt added over 1m points to my total yet.. overclock site is good though
idk how to check progress in anything other than the web client
find your username on the EOC website
F@H stats folding teams overall rankings. Help Folding at Home fight Coronavirus, further medical research, and prevent diseases with distributed computing!
its not perfect but its the closest you're going to get
If you are like me (Likes to watch your pc working) When you current job its a 95% on GPU/CPU just create another slot with 4 thread and F@H will start looking for a job for it, when your other job finish, just delete the 4 thread Slot and The WU will be moved to the regular Slot. And you don't have to wait for it to send and the Download a job. It will start working Immediately and send it at the same time. Saves time and get WAY higher PPD 
there is a config option for that Slid
Its not that eficient I had it on
you dont need to do all that rigmarole
stats for my pool haven't been updated in 3 days. is this normal?
250658
alright that bsod me yikes
what did?
splitting the cpu
yeah once i reopened it crashed\
is it OC'd?
its stock rn
huh
maybe due to the other stuff im doing rn
no longer open in task hmm gotta find it again
thats weird. we had stats until like 3 days ago
Close anything on TM called Fah
is it one of the teams that weren't making points that got removed?
ohh
And then re open
theres a write up on it on the main page
"Small update for Sunday.03.29.20, 2:26pm CDT
Before I get too many emails asking why there are negative WUs in the "aggregate stats", there is a simply answer to that. I'm tracking the top 12,000 teams"
oh wait. our team is too small too new to be on eoc
Latest news for EOC F@H stats. Help Folding at Home fight Coronavirus, further medical research, and prevent diseases with distributed computing!
the EOC website still says I haven't completed any WUs in 5 days even tho I know I did 2 yesterday. I'm gonna keep folding and see if they show up eventually
main website updates slow
ah alright
ask your buddies to pool money and build a server for folding
then get your team in the top 12,000
WIN
π
yeah we're pretty broke though
sure feel free if you want to
After I get chu like 1m I will change back
you can fold for ltt though its almost # 1
I dont fold for LLT because they dont need the points anymore. I will always fold for a cure yes, but give the points to smol teams that need it π
I see. thats really cool of yo
I think i've only done like ~80,000 so far
do you have your passkey? make sure you get one to get bonus points..
I would fold for a day but right now im competing with a friend of mine in the LTT Team
also after your 10th verified WU you get a quick return bonus (QRB)
basically its just a boost to your PPD
but you also need the Passkey
guess ill wait for the key
make sure your username is the same for both
it may take a few hours to get it.. just keep checking your email you sent it to
i've already broken my record for actually daily PPD
and we still have 6 hours to go
thats funny
~20 years ago I worked in a computer store. Customer told us what spec and we built it. I remember building a "killer gaming rig" for this family with a teenage son. 6 months later the mother comes back, with the computer, furious. She starts yelling before I even get to say hello.
The computer reboots/turns off while the son is gaming.
I calmly let her know that "since it is within the first year I will look at, and if possible, have it fixed for them by tomorrow morning". She leaves. I open it up and immediately close it again. Air on can won't get the job done. I have to use the compressor. Holy crap that thing had been collecting dust.
I clean it, replace the cpu thermal paste. When I start it up I notice that the multiplier is not what it's supposed to be. Turns out the little shit had placed the machine in the dustiest basement corner AND overclocked it.
Forum posts like the one above reminds me of that. People...
Whats airflow?
65 C is not hot so you need to fix your pc.
It is if youβre running a watercooling loop of a noctua at stock speeds

What were they expecting with an i3
max temp is like 72c
Okay, and that next thing about that kid though
I would have been like "what the shit were you expecting the dust to do?"
two words: fan filters
lol, yup
I'm actually really happy because I've been getting WU's all day today for my CPU
that's awesome
I got permission from the client at one of my sites to run F@H on their Quadro P6000 / Xeon Plat. 8180
sick
Can't wait to get that set up eventually
GPU Temps never go over 60c
i mean we haev all our client sites on anydesk
i really doubt my boss would go for that though lol
It's funny to me because the client came up to me and asked if it were possible
I was like "wait... you know what that is?!"
that's a MONSTER
I should mention
all.the.folds.
He's got about 4 of those rigs in rackmounts
I honestly don't feel much of a difference
I know for sure a SSD makes more of a difference over a hard drive, but as for ram, not sure
most of mine are on emc sasflash or ssdflash
VM's right?
You should be pefrectly fine. How many cores allocated per VM?
4 per
yeah, you're golden
yea i'm averaging 100-120 wu's a day
Doesn't look like it allocates much to RAM when running the calculations
my free -h shows only like 570mb allocated
If you have nothing else running in the background I'd say you're good
If you want to err on the side of caution you can add 512, but I think you're fine
I GOT A WU! Finally.
there's actually a full prod environment running on all sites, but on a evc cluster. i'm running on seperate hardware
nice jo!
Pre SARS-CoV-2 research: Not long
Now: Don't hold your breath
usually every hour
Am I missing it or is there no way to fold specifically for Covid-19?
Any Disease = COVID
I thought so, thank you
They were smart about allocation of resources by making it "any disease" tbh
linus pls
https://apps.foldingathome.org/serverstats
this entire thing about getting a whole bunch of LTT folders and new ones is about pooling our processing power, mainly GPUs and the LTT server isn't assigning GPU jobs
That way when all the covid WU's are taken up, we can at least fold for something else in the mean time
Anyone know what the GRO_A7 is?
gro_a7 i believe are CPU jobs
and openmm_22 are GPU jobs
probably 21 too but idk about those, those i think are experimentals
at least that's my understanding of it
21 is the previous GPU core. 22 is the current.
wrong link
The distributed-computing project Folding@home uses scientific computer programs, referred to as "cores" or "fahcores", to perform calculations. Folding@home's cores are based on modified and optimized versions of molecular simulation programs for calculation, including TINKER...
ah
So switching to linux would help them out more?
That's interesting to know
I should try to set up the the remote access to other clients
just haven't sit down to read it yet, but would be nice to reset the pc sitting at work remotely
223518
That is not
That is not what?
Oh, it is not LTT you are after?
no
@trail night 250658, but 223518 is the better team π
Thanks
There is no better teams, just one bigger than other. but we all contributing for the same reason.
is the number 1 team effectively paying people to fold?
Yes
But for most it won't pay for the power used to fold so it's pointless to fold for them
ltt already #1?
well, if you're folding anyway, something is ebtter than nothing. That said, I feel like folding is more of a charity thing in spirit, so carry on with team LTT.
EVGA is better to fold for if you want money, you need to manually claim each month to get the store credit but it is actually worth it if you remember to claim every month
then you can pay down a new GPU and fold for LTT during events
Interesting. Might explore that post covid situation
wait what you can get store credit for folding for evga
Yes but read the terms carefully if you want to go down that path, it's either $5/m or $10/m depending if you meet the minimums and remember to claim
30 is $120
max for the year
Yeah
in fahcontrol, right click the slot and hit finish
in fahcontrol, right click the slot and hit finish
@weary epoch hes asking in boinc, not fah
ah oops
π
@vast shell "no new tasks" or similar, when in project view
so for max EVGA bucks you need to fold about 1 mil per day corrrect?
yeah
struggling to get 1 mil/week with the lack of WUs, despite having potential for 2 mil/day in theory
(granted I don't run 24/7 anyway so even with WU's I wouldnt get that in practice.)
desktop says PPD is ~850k but really it's probably closer to like 300k because of lack of WU
my points for today is 830,482
my 24hr avg is 746,914
laptop
I got lucky today with WU so I'm ~500k, running only on my 1080
nice
my computer has truble getting WUs at night but during the day its loaded up
ah
@median vine you want my Actual or Estimated PPD?
eh it's fine
ok
did the math. i think it would have took me 10-14 days of folding nonstop for EVGA to get $10.
and then you got to redeem it every month by logging in or its forfeit
if i had more folding capability it would onky take 5-6 with better hardware. still its on the folder to redeem it for store exclusive stuff.
its useful if you are a regular @ EVGA
i think you can use evgabucks to purchase an extended waranty on an evga card you just bought though. dont quote me on that.
folding for a year to get an extra 5 year warranty.
the way i read the member club was that the evga bucks are good as cash when you check out? i could be wrong
@stark bough The other thing is EVGABucks have to be supported in your region, it is not for me
Main rig has been folding every night for about a week now. Tomorrow I am going to get some of my old computer hooked up and updated and installing Folding@rancid agate to get more points a month.
@unique sigil When is the LTT FAH server video going up, dying of anticipation
thought linus said it was monday for floatplane and maybe a few days after for main channel or they might do it at the same time
looks like the LTT folding server is down
is it normal that folding only takes ~40% of GPU power?
I thought he said this weekend, but depends on getting that last bit of footage
yes. programs misreport it. if you go to cuda, it will show 100%
Nice!
its not down just out of jobs atm
ahh
Level 2 cache finally almost full
π
already 500GB of data stored on the server oof
you need to tell them that they need more work for higher core count CPUs. when set at 16 threads i got almost no work split it down to 3 sets of 4 and a 2 and non stop WUs
@unique sigil good job Jake!
lead do you know if it's normal that if set for 4 threads it loads up all 16 at 25%. instead of 4 at 100%?
I'd be interested in the typical I/O block size, IOPs, sustained bandwidth, CPU load and memory but I expect like 2 or 3 of those in the video
No it should load 4 at 100%
it's not
Have you set it at light workload? Or full?
full
now that its been running for a bit sustained inbound is like 20-30megabit
Thats not much
that'll be 100% the WU shortage
Seeing as you have a 5gbit pipe to that server right?
no jobs, no bandwidth usage
ya
The NVMe cache is putting in work
average utilization is like 30-60%
actual HDD usage is below 0.3% on each
On each? How many servers did you put up?
each HDD drive
Oh
That usage percentage is capacity used in the cache correct? Not like performance utilization?
The NVMe cache is putting in work
average utilization is like 30-60%
that?
read/write utilization, not capacity
theres current usage of ARC + L2ARC cache
hmm, higher than I was expecting
not sure how familiar you are with ZFS
but basically L2ARC works by taking anything that is being removed from ARC (level 1 sys memory cache), and rather than putting it back on the base disks, it throws it on the L2ARC cache.
no gpu task available for me, for the last 11 tries. bad ltt server π¦ ( just kidding)
Enough, more just I expected better from NVMe honestly
LTT server has not even had a GPU job to hand out yet
I mean its reading from system memory which is real quick
So it actually seems pretty reasonable to me
optane 905P so not that craziest throughput but really gnarly IOPS
yea, well I mean the total throughput is kinda slow but then I don't know the IOPs so I'm 100% just speculating on like not enough info
network throughput?
i should have 6 thread open/ideal jobs.
From what I understand they just need to add more jobs to it
Yea, but is the server also doing other processing to create/process WUs which cause storage I/O?
- there is a default cap on network connections
they're gonna bump the network connection cap tmr
I'm not sure actually
π
how much are you actually doing for the setup btw Jake? do you configure it? or just let them remote into the server and make sure its online?
Probably hardware build, rack, network connectivity, OS install, SSH access setup
also, awesome that you are just randomly in the chat answering questions/talking about this stuff! very nice that you actually do that stuff
^
@rich depot Particularly my super nosy IT questions π
I don't think they are nosy, we are all pretty interested in IT so it makes sense to wonder about this stuff
Yep, but I always want to know the detailed stuff that is never in the videos
Mind you that nobody really want to know either so doesn't belong in them
the dirty work that goes into it
Mind you that nobody really want to know either so doesn't belong in them
@silver wing
Just make another channel and throw the techs stuff in them

^ or detailed writeups, if they dont already
Ain't nobody got time fo dat
Inb4 30 post long write up on twitter
So what kind of server is LTT deploying?
Distributing CPU workloads?
new channel that just randomly walks around the warehouse to see what people are doing
It's the one for folding at home
yeah, thats the downside of making video's for 10mil people, you can't go hardcore nitty gritty into the technical stuff. But that would be an awesome new channel, like real hardcore IT stuff.
π
that it would
Nah, GN is server newb
level1
I have a theory.
Jay, Kyle, and Paul arenβt actually in Southern California. They all actually live in the same studio backlot in a common building behind LTT in Vancouver.
na Jay has his own building
Even L1 kinda limit themselves to subset of things, but yea more server and network stuff but nothing I actually work with
Oh yeah thats awesome
Old man yells at Standoffs
It's the wrong hole!
Picture above is of someone trying to place a True E-ATX motherboard in to a case that supports Full E-ATX
Mainstream E-ATX
I got a true E-ATM in a midi case once.... Bring it on!
Trying to place the 5th RAM stick in the DIMM.2 slot on an Asus board
eGirl ATX
PS there was a Dremel and lots of booze involved
Well I stick to SSI-CEB motherboards in towers so there
Trying to place the 5th RAM stick in the DIMM.2 slot on an Asus board
@fierce steppe but doing it with a fake stick for only the extra RGB!
Just use optane as RAM
That the same damned thing!!!
Well, no it's not but π
8 GB RAM + 16 GB Optane = 24 GB Memory
It's pretty the same thing
Ouch
CEB doesn't have 4000 different sizes
@stark bough if they have RAM listed, whats memory?
Name Form factor (Height Γ Length)
Compact Electronics Bay (CEB) 12 Γ 10.5 in (305 Γ 267 mm)
Enterprise Electronics Bay (EEB) 12 Γ 13 in (305 Γ 330 mm)
Midrange Electronics Bay (MEB) 16.2 Γ 13 in (411 Γ 330 mm)
real standards, yay!
Itβs memory
Optane plus RAM
Soon= βdo I still need to buy an SSD and DDR4? Canβt I just buy optane?β
Thought it was a myth
woops mine are actually EEB, oh well
Why? Not like Ryzen has the PCIe lanes to bother
I have trouble getting WU... seems like we sucked the queue empty
I got 2 million ppd yesterday anyone know the reason for the sudden drop?
pc crashed due to memory being too hot and I lost my cpu WU 
Can you see gpu memory temps anywhere? Getting 51 degrees in afterburner
only if there are thermistors on your memory.
@safe horizon you know when sharing a 2 screen screenshot nobody has any idea where to look for what you mean. or probably can't see cause the screenshot won't open in 2 screens
@stark bough Would that make sense though if I'm only at about 50 degrees in afterburner?
My Server has only one slot and has two WUs ... 
it tries to download one just before the first is done
it tries to download one just before the first is done
@safe prairie its normal for fah to do that, since this is just looking for new wu to run straight after the previous wu done
does 10 days active mean 10 days with a submitted WU or 10 days running?
10 days with a submitted WU, out of the 14
@safe prairie
Had that error before aswell but restarting it a few times and so fixes it.
my brain feel smol
@safe horizon I got ~1m points 2 days ago, but only got 90k yesterday due to low number of WUs, but it seems to be back to normal, for me today
hey @rich depot do we have a update for yesterday?
It's already up
ah found it. was looking for a fourm post did not look in the event post
woot up almost 200 spots
spectrum is it possible to add a #change column to the sheet?
The last few days I got like 100-150k points
Should have gotten at least 7x that at minimum
Let's see if it improves today
@supple laurel might wanna tag @rich depot for your question :)
thank you @silver wing
can someone link current stats page?
For?
the answer
10 days with a submitted WU, out of the 14
@silver wing
That's one of the WUs from the LTT server
F@H stats team summary for PC Master Race - PCMR. Help Folding at Home fight Coronavirus, further medical research, and prevent diseases with distributed computing!
mine is meh
is there a way to translate the timeline on EO from what looks like east coast time to other timezones?
Yes, the same way you would convert time zones normally

I don't fully trust eoc though
If you look at ltt there have been several times it was 0 points for that update, but also the one later wasn't double or something
That isn't EOC that's the FAH export not updating so the data file is the same
No update to data file no increase in points so 0 point update on EOC
When you download the data file from FAH the first line is the date/time when the export was created
Am I correct that Amazon's EC2 Spot instances are making use of idle time on their servers?
With EC2Spot becoming a threat to the LTT team
Could we just buy up some time on the servers and have it fold for the LTT team, and then that would also stop the EC2Spot from having idle time to run? π€
is the extremeoverclock website region banned?
i cant access it from all devices that i have
all result in server timed out error type
With EC2Spot becoming a threat to the LTT team
Could we just buy up some time on the servers and have it fold for the LTT team, and then that would also stop the EC2Spot from having idle time to run? π€
@fierce steppe big brain strat?
π€
estimated credit jumps 10k just from that
sometimes fah core unit got special treatment from windows for a short period of time tho
i came across the same behavior like yours (for one frame only)
1 min 40s was what I expected tho
i check my control now and then and it's around that number on average
left it overnight, then when I checked again it was 2 min 30
and gpu util was all over the place
usually would've been constant 90% or so
you prob tur off your monitor when running through the night right?
ooo that might be it
ooo that might be it
@marble quartz the monitor got turned off?
btw thats the cuda utilization right?
@trail night 250658 is the code you were looking for
and damn. was that you? someone doubled our points overnight
it look us a whole week to earn that amount
nope that's not it
nvm
well compute utilization. vega 56
π
yknow, turn off your monitor to reduce the power usage of it since it will consume more electricity than your pc in some cases
Yes itβs me @rancid flume
yay welcome to the ltt team π
590 more than 0
Wait no
I didnβt got jobs overnight
It was not me
I donβt think my other 3 pcβs were running
folding stats slow to updates?
i cant access eoc website somehow 
also lets you see hourly/daily/weekly stats
eh it's fine on me
Latest news for EOC F@H stats. Help Folding at Home fight Coronavirus, further medical research, and prevent diseases with distributed computing!
well i fooked up my gpu folding with an unstable oc and now its sending half completed stuff
...
ERR_CONNECTION_TIMED_OUT
got this error all the time
i havent try changing the dns yet (btw im using cloudflare 1.1.1.1)
usually it dumps the job for me if it failed
oof
one time i finished a bugged wu in 3 days
when it sent the server rejected it and all my cpu hard work went down the trash can
F
After all these years I guess I'm finally starting to put all the Folding@Home data I've been collecting to good use! This is my first attempt at creating a chart race using points from the top teams from July 1, 2008 thru March 29, 2020. Yes, I know I made the end date in the...
i like the music
heres what i got changing the ipv4 from cloudfare 1.1.1.1 to googledns
any ideas to fix this lads?
LTT is lower than the default folding team again.
btw this currently only happens to any extreme overclocking website
refresh DNS?
i had issues with 1.1.1.1
I use 8.8.8.8
refresh DNS?
@stark bough flush it with cmd?
the global mining stats on the folding@rancid agate official website is like 3 days lagged atm
dunno if you can just punch in the ip address of https://folding.extremeoverclocking.com/ into your address bar to get it
Latest news for EOC F@H stats. Help Folding at Home fight Coronavirus, further medical research, and prevent diseases with distributed computing!
someone just added a shitload of processing
meanwhile did PCMR just pull a ton of cpus/gpu for game releases?
@rotund flume ipconfig /flushdns
my TPF is around double what it should be; what's going on
different projects might have different TPF
yup
damn, aws is doing great with folding
I googled but google didn't know aynthing
yeah i've pinged them on twitter but no response yet.
I have gotten 250k points today, twice that of I have usually doe the last week
But still not close to what my computer is capable of
AWS has to 10x what they're doing to be any kind of threat as we're still significantly out producing them and they have to not just out produce us but by a lot to catch any time soon, and they will lose interest
But I think they will crack the top 10
true, but they are rising fast
Is that jump in daily points LTT server going online?
Nope, it's not really doing much, it's only 1 server out of 21
and from what jake was saying last night. it's not really up and fully running yet. it was kind of in test mode over the weekend.
I'm sure if they open the flood gate on it it'll run up the WU/hr it's doing by a lot
Assuming it's got a decent amount of ram and good cpus it should be possible to out do fah1.eastus.cloudapp.azure.com
More servers is definitely still needed
the ltt server is semi up, nvve was up to 60% usage I believe. lots of iops
checks LTT server status
Assign mode 34 work units
refreshes page
Accept mode 0 work units
@rich depot https://apps.foldingathome.org/serverstats
yeah but here it says its assigning 8k wu's an hour
that's how averages work
so where are you looking spotty?
You can assign 200,000 work units in one hour and then nothing for the rest of the day and that would be an 8k/hour average
ah okay, I just look at the first number
didn't know the last 2 were the actual that are now out by that server
I've looked but I still can't find a write-up on how to read that page π¦ lots of people have speculated when I asked, but no real answer yet
Jobs columns is the number of jobs the server has to give out right then, if it's zero then it's got none left to give
^
ah okay ty
I hope they have let it run out so they could change a few things and make it a more effective server
Assign = it has work units to distribute and is handing them out to people. It will also collect any of the work units it distributed that have been completed
Accept = it doesn't have any work units to distribute but the server is still online and collecting the completed work units it sent out
Down = offline, can't be reached
thanks
and has cs? and is there a difference between public and beta jobs? cause I see a shitload of beta stuff
Change what to make it more effective?
As far as I can tell it's able to distribute all of its work units very quickly...
CS = Collection Server
maybe have it produce more wu's, or distribute them even faster
CS = Collection Server
@fierce steppe
Called It!
Called what? If you hover over it it says "no collection server"
The top two servers send WUs to the WS servers btw
i speculated on what everything on the server stats page means a couple days ago
ah ty @silver wing
Each project is assigned a work unit that will distribute it
Since the LTT server is new they have to wait until they create new projects to assign to the server
they can't put existing work units from other servers on to the LTT one
makes sense
not to me it doesn't but what do I know π
16404, 16406, 16408, and 16410 are the ones currently assigned to the LTT server
seems like a bad scaling limitation but scaling wasn't required before lol
yeah it wasn't really an issue until a few weeks ago
LTT should do a technical video on how this all works and how they kept up with demand after this is all done and working stable
So each project is coded with a work server and most will have a backup "Collection Server" if the work server is unavailable
When a work unit is completed and gets uploaded, if it can't connect to the work server it goes to the collection server.
The collection server doesn't have the information on when the work unit was sent out from the work server, so it can't calculate the QRB so it'll only show the base credit in the logs as the estimate and it won't credit any points until it they get uploaded to the work server.
The collection server does track when it was uploaded to it, so it will give that information to the work server and the work server calculates the QRB from that, so you still get the QRB
that would explain the delay on WU credit in some cases
and that is also why some people were having work units that couldn't be uploaded for days at a time when the work server & collection server the projects had assigned were down
9.6.1 I think has a fix for the QRB so you get it always
do you have a source for that?
I've had several work units fail to credit today for going to the collection server
Haven't checked to see if they were uploaded to a CS running 9.6.1 though
That was a change they were making and 9.6.1 is a release after that, but no guarantee it's actually in that update
they are pushing out new software versions very fast though there are 4 different versions running atm
Yeah it was like 9.5.2 or something a month ago
yeah, there are now servers with 9.5.2 - 9.5.8 - 9.6 - 9.6.1
oh also a 9.5.6 so 5 different versions
reading the folding forum makes you appreciate how clean the LTT forum looks
I don't visit the ltt forum
I tracked this down. Sadly WUs returned to a Collection Server (CS) are not getting credited immediately. This is likely been going on for a long time. Normally, WUs would get forwarded to the WS fairly quickly and get credited quickly. Since we are under such huge loads the CS have backlogged and many accepted WUs are sitting there uncredited. All these WUs will get credited eventually. Bonuses are calculated from the time the WU was returned to the CS so bonuses should not be affected. I'm pushing a software update which will cause the CS to credit WUs immediately in the future. Thanks for providing all the info to diagnose this.
The last one I did a few hours ago that failed to credit was to a server running 9.6
I'll have to keep an eye out to see if it's fixed on servers running 9.6.1
Damn it, I only just found that post and was just about to link it lol
Checking my logs I did have one go to a collection server that was 9.6.1 and that was credited normally, so perhaps 9.6.1 did fix it
If they can get that fixed and back credit the missing WUs it'll give a huge boost to our daily points
Roughly half my WU's have failed to credit because of that
Yea 9.6.1 started popping up shortly after that so I used my powers of assumption π
Missing QRB has been a problem though, I'm sure of it
That's yours?
atleast your getting some im only getting WU's every other 3hrs π
If they go to the collection server it will show the base credit without QRB in the logs, but you won't get credited for the work unit at all as far as F@H is concerned - at least not until they sync up the collection server to the work server and import the jobs over
it won't show as completing a WU in your stats
But
Since you're doing mostly CPU jobs
It's possible that you just got a bunch of jobs that were really short and had a low credit
CPU jobs vary from 140 base credit to 5,500 base credit depending on their size
I should be around 4.5m-5m PPD so each update of mine should be 600k +/- 50K
you probably are missing WUs that aren't being credited
But it shouldn't credit you just the base credit without QRB
Dunno, most of my systems are unmonitored and CPU only
but I know for sure QRB is a problem again
Yeah I'm expecting to get like 5 million points one update when they sync it up
rookie numbers π
shhh it's only been happening last 3 days :p
well at least that I've been keeping track of
randomly get 100m one day in to a folding event with no explanation lol
Oh by the way I told everyone you were racing TVwzhere to 500m
loser has to mod the discord π
starts up 20 more servers
If only F@H would do that
They're actually coping relatively well considering they went from 30k users to 700k users in < a month
looking at P3 instances
honestly I think the WU rate is at what is needed now, just need to get the stability back up
oh welp the spot price...wow
100k WU/hr and EOC has 100k WUs been done every 3 hours
Something I don't understand with the F@H stats reporting for work units distributed
46k GPU work units
40k CPU work units
total = 109k
confused math lady
rounding, has to be rounding
Ah they must be using the Sans rounding formula
they do say the assign rate is estimated
goes down 18000, shows 9000
but I would think the total would just be CPU + GPU = Total
They must be calculating the total rate off something else
It'll be interesting to see the EOC graphs update at the end of the month
Our previous best month was last years folding event at 15.5B points
We've already done 29B points. We'll easily double it
68b
If we hit 100B for the month of April Twitter will add an edit button
CEO of twitter told me himself π
so never
mind you a lot of the folders are not part of the folding event so maybe they won't stop
event only covers 8000 of the 60000
π€
I think there's about 2500 in the event who haven't submitted a WU yet
If they haven't submitted a work unit at all they won't be counted as a user
Some people either signed up before the event and got frustrated that it wasn't working and quit, or they forgot to turn it on, or they couldn't get it to work
then there's always people who just sign up thinking they will get free prizes
Or maybe they are just using a Pentium 4 processor and it takes a week to complete a work unit
They're trying their best, dammit!
Would leaving my Ryzen 5 3600 folding 24/7 at near 90 degrees be a bad idea?
I had a WU on my laptop take around 38 hours
@mild dawn thats a little warm, personally for 24/7 I like 85C or lower
90C is a bit on the high side
Increase the fan RPM, take the side panel off the case
386DX master race!
nice just got CPU WU's on 4 different machines
I still have no idea if there's anything wrong with my cooler or it's just how Ryzen works
I'm not a Ryzen Expert, but I think most motherboards give a little too much voltage stock
BIOS update might fix that
or undervolt yourself
if I were you I'd do both
- hang on i actually have a 3600 folding right now
let me see the temps
Already updated BIOS and toyed around undervolting/overclocking but temp didn't really change much so I reverted to stock
do your temps drop with the side panel off?
Ryzen Master is currently reporting 87 degrees and using 1.32 volt
It dropped to like 85
stock cooler or something better?
Me
Currently running a deepcool gammaxx gt
69C at 1.39V
Not the best but better than stock
Noctua NH-D9DX i4 3U cpu cooler
Well we know why yours is 69C π
After a while temps are 85 now with side panel off, yay
Maybe check mounting and paste, maybe you can get 5C lower than that but I don't really know
I did that 3 times π¦ the first time actually made it run cooler by 2 degrees
Maybe it's just how southeast asia climate works
can you set the cpu fan to 'turbo speed' i.e 100% and see if the temps drop a lot
Get one of these π
full custom EK loop nice
386DX master race!
@silver wing 486DX-2 laughs in FPU π
might need a top up there bud
I win
Those are the dual 290Xs?
What case is that?
I tried watercooling once, it didn't go well
LD PC-V8
loop is still in emergency repair for like a year
it WAS all hardline lol
There's like 5L in that loop total
taken it to one lan event
never again
very sure one of my PCIe slots is damaged
i joined and started folding this morning
Aiye welcome to the team
GPU sometimes doesn't work or won't crossfire, it's the reason the loop is all apart and random tubes. Hardline = impossible to troubleshoot
It certainly opens Microsoft Word 10% faster, so RGB is worth it
getting my money's worth out of this Corsair Obsidian 650D case
tbh I like more of the older cases than the new ones
HDD bays are just not a thing anymore
my case came with a 4 bay and 3 bay, takes 7 total with 2 2.5" mounts on the back of the motherboard tray
but if you have all the HDD bays in it blocks the front intakes
not any worse than a glass front does
put this in the 5.25 bays @fierce steppe π
its pure base 800 so its already pretty restricted
you'd have a NAS at that point lol
I have NAS hdds, does that count?
if more than 3 and zfs than yeah π
hehe same here I've got a pile of old hdds 1 and 2 TB all nearing their bad sector death and collecting dust now
Damn, Bannerlord is $70 AUD
reminds me of the pile in our maintenance room where we upgrade/test the EMC boxes π
the 3.5" drives are 3TB each, don't remember how many but I pulled them out of a 48 bay shelf
I have 3 more HDD's downstairs
2.5" are 900GB but you can't see those
what's the highest hour HDD you still run?
hour? Like total power on time?
yeah power on hours
no idea, ages
run em till dead
24,122 hours on my games SSD
yep, let them run until they click π
most of my stuff is 24/7
followed by 11,000 hours on my HDDs
I know I have stuff with over 70,000
those 3TB all came from a system that was 5 years old 24/7 so they'll be in the 40,000
ohh i have an 840 pro in another machine
let me get the hours
110 hours difference LOL
haha must have got them about the same time
03/31/2013
oh wow
literally 1 day away from turning 7
well 7 years from purchase...
I paid $140 for it
No idea what I paid for it
well back then around 1$ per GB was common
I've also got 6 512GB 850 Pro
Just use the price history from honeyπ
Got the older 840 Evo, I think 2, I know one of them died
like I went to put it in a laptop, didn't show up in Windows installer so I ignored it for a year
got back to it, confirmed it was dead, 1 month are warranty
Morning FOlders
that annoyed me a lot
How have you done since midnight Gansta?
Nice gangsta
I wasn't dead, it was last used in Storage Spaces as a cache/tier SSD and windows installer ignores disks with Storage Spaces metadata on them
makes 100% sense now
That is why I generally only buy Samsung
Made the mistake of buying 2 Sandisk SSD Plus on the cheap
they died
Like 7?
same with my Sandisks
ohh i guess i've seen 2 SSD's die then
for me Samsung is the only producer of SSD's
My 3600 reaches 70c max
Be Quiet Dark Rock 4
I know it's 24 of the 2.5"
and 9 Velociraptors
ohh
Yea those 3.5" converters were originally in a StorSimple and had SAS SSDs in them
My NAS is 2x8TB
At least yours are in a NAS π
One I built myself, unRAID
https://sgcdn.startech.com/005329/media/products/gallery_large/25sat35hdd.b.jpg
You thought I was a 3.5" but suprise, its me, a 2.5"
My NAS is very small but can still fit up to 7 3,5"
cool though
Lian Li PC-Q25
rip maxtor
to the MAXXXXX
I don't even bother with my old stuff like that
like my 80GB ones
still got em tho
you see the seagate laptop drive?
that actually came out of my Original PS3
LOL
i needed a boot drive for FreeNAS
actually i think its still in there...
i put a 160GB WD in the PS3
forgiven π
just shove a fattydove in there
oh great now my 840 Evo is constantly disconnecting and connecting
windows is having a spaz
Yep!
only has 25k hours on it
though thats probably alot for a "Laptop" HDD
@silver wing you have any IDE drives?
a 2.5" laptop
jinx
see if I can find a pic of the laptop we played scorched earth on at school
Are you guys digging through the lower layers in your closets hunting for folding hardware? o.0
Be Quiet Dark Rock 4
@pseudo abyss I want one of those or a similar Noctua
@smoky rose I just put my ibm compatible to work... So maybe
ask me about that when I was 7 and id think it had its own printer inside.
wonder what the PPD is
1 Point Per Year maybe
Nopes, deadlines will kick in before that.
lol
would have to be zero then
@fierce steppe The total assign rate is the sum of the two AS servers
realistically probably ~200-300 points per day
linus1 has been sitting with 0 WU for a while. Now it's up to 2. π
My SSD for old desktop is 26537 power on hours
@fierce steppe The total assign rate is the sum of the two AS servers
@silver wing Ah... Still doesn't explain the discrepancy between the project assign rates and the server assign rates
Hey guys just had a question, I've been folding non stop for about two weeks, and like 3-4 days ago my stats page stopped updating, the web interface registers 1.72m points, but the stat page has been stuck at 1.07m points for those 3-4 days, any idea?
Lolz I read the title to this discord section wrong xD my bad
The FAH servers are "busy" doing other things. Stats will be updated eventually.
So, how long does it take for the official website to update your score? It's been sitting at this number for 3 days now. My web control panel is accurate, I'm guessing...
THE pioneer of distributed computing and boinc, SETI@home the inventor of grid computing is going to hibernation and shutting down its work distribution tomorrow on 31st π maybe its worth mention somewhere
Don't bother with the official website. Go check on https://folding.extremeoverclocking.com/
Latest news for EOC F@H stats. Help Folding at Home fight Coronavirus, further medical research, and prevent diseases with distributed computing!
It tracks the top 12000 teams.
@near gazelle Your stats: https://folding.extremeoverclocking.com/user_summary.php?s=&u=953874
F@H stats user summary for zombienerd. Help Folding at Home fight Coronavirus, further medical research, and prevent diseases with distributed computing!
Thanks
How can they pull this data, but the official site can't do it properly? Lol
benchmarking with f@h
Right now I'm just sad I missed the conversations about the vintage laptops π I could have uploaded pictures of a few of mine lol
i closed my FAH process like a moron and had to start my 7-hour WU from the beginning
My oldest one is from 1987 - Zenith ZFL-181 - there's a video about it on my channel, the pop-up floppy drives are so much fun lol.
@near gazelle The data is available for download by anyone. The official status database can't handle the load.
My impression is that downloading has improved somewhat but reduced congestion there (if that's what happened) has allowed us to see congestion on back-end systems ... like the points DB.```
https://foldingforum.org/viewtopic.php?p=320254#p320254
I did get a response from the F@H team RE: spinning up new WU servers. They require 40TB of storage, and at least a 1gbps connection.
Looked into dedicated hosting, probably around $250-300/mo with a bandwidth cap. Couldn't find any servers that were unlimited bandwidth that had that kind of storage on a gigabit connection.
I can get 1gbps in my home..
https://apps.foldingathome.org/serverstats
There are servers that have not been fully deployed yet.
what practical purpose is there to a bandwidth cap
i never understood that practice
I refuse to go with an ISP that limits my ability to use the service
even datacentres have a limited connection. so if you have a bandwith cap, you probably won't use your full gbit 24/7. leaving room for other people
well, yestergear was looking at shared hosting so thats that
I would be wary about folding/mining on shared hosts. Some prevent it in their AuP
i dropped the CPU slots, the 1060 limited to 60W is quiet and still does some work...
also, most home connections are overbooked, so they sell you a max speed, which you probably won't get at the peak hours. so setting a bandwith cap reduces that too. although that practice is extremely bad and stupid
@rich depot i am lucky, my ISP is not overbooked, nor data capped π
I know of datacenters with unlimited bandwidth, but they don't offer 40 terabyte arrays lol
fortunately here they have seperate fiberlines to each home here π
Just move to a developed country and you will not see that practice. π
or you get some shady isp who offers you "gigabit ftth" and then only guarantees gigabit speed TO THEIR own router
like canada @smoky rose ?
good enough for folding π
Northern Europe.
While I'm sure they'd limit F@H number-crunching, I don't think they'd have an issue with file / workunit serving
I know its decent there (I'm in the netherlands) but canada still sucks appareantly on that front
can you get a connection that is bandwith capped?
I live in the middle of nowhere in the USA, best I can get is a 100mbps down / 20mbps up connection.
cause I literally can't get a bandwith capped connection at my home. On my phone yeah (although there are unlimited plans) but on my home fiber connection, nope
nice
300/20
I don't know how much I use. I do know that when the archive team was trying to archive google+ I used up my 500 mbit/s for like a week straight, before google got annoying.
i have my QoS capped at 80% of my total speed
that drops my bufferbloat to almost nil
I believe I'm back to 100/100 right now though, I don't really need more usually, but I can upscale to 1gbit/s within 24 hours I believe, and downscale after 30 days to prevent paying to much
as long as you have a prosumer router not some shitty walmart brand
I use the router the internet company provided and it seems to work great π I believe I can run a custom router but no need
hmm
I usually leave those in the box
You can get an ER-X for like 50USD they perform like no one's business
is there something wrong with the folding at home laeaderboards? my points havent updated in days
yeah, but why would I spend 50 usd on something I don't need?
despite the webpages almost quadrupaling
@devout wing yes, they can't keep up
you wouldn't lol
F@H stats team summary for LinusTechTips_Team. Help Folding at Home fight Coronavirus, further medical research, and prevent diseases with distributed computing!
ok, so as long as I see them on my web dashboard im getting them?
I use a Tenda AC router. Had it for about a year and a a half now. For less than $50, I was insanely surprised at the throughput.
yes
there can be a delay on getting points, but in the end you will get them all π
I seperate my wifi and router. i got one of those ubiqiti ufo dealies
AP-LR... I can get wifi for like half a kilometre
next target is Team Default π
yeah, your neighbours must be really happy with that π there is like a few hundred homes within half a km from me
no one around me is on 5ghz for some reason
you do half a km on 5ghz?
yea
In the 24 hour timeframe, we're already number 1
It gets spotty out around the edges and without a really good line of sight, but I can go to the bus stop about 300m away and still have signal
yeah @mystic hollow
What I find really neat is that in just 3 months, we are almost passing the entire WU load from the entire 2019
we can beat that easily
If the FAH servers start handing out WU without delays and we keep folding and growing we will overtake them all.
i'm going to the office today to spin up another esxi host
Yeah, just waiting on more workunit servers lol
yeah, the more wu's, the faster we overtake them
where are all the WUs
although I really hope the covid crisis is gone before we overtake curecoin
If a few more colleges could spin up some WU servers, this shiz would get crazy quick.
@storm talon how many servers are you running lol
Hah, the competition only requires 500k? I thought it was 5 million when I looked the first time lol..
I was worried I wouldn't get there, but already tripled it.
3 datacenters with 2 clusters of r810's in each
I can only use our main office for folding though
and how many is that?




