LTT Folding Team's Emergency Response to Covid-19

[GOTSpectrum]

7 minutes ago, TrueMad said:

nobody can stop me ... but i'm wrong...

 

Collecting server answers: IN YOUR FACE

19:09:32:WU02:FS01:Uploading 23.06MiB to 155.247.166.219
19:09:32:WU02:FS01:Connecting to 155.247.166.219:8080
19:09:38:WU02:FS01:Upload 13.82%
19:09:44:WU02:FS01:Upload 28.18%
19:09:50:WU02:FS01:Upload 42.82%
19:09:56:WU02:FS01:Upload 53.11%
19:10:02:WU02:FS01:Upload 68.02%
19:10:03:WU00:FS00:0xa7:Completed 70000 out of 250000 steps (28%)
19:10:08:WU02:FS01:Upload 82.38%
19:10:14:WU02:FS01:Upload 97.01%
19:10:15:WU02:FS01:Upload complete
19:10:15:WU02:FS01:Server responded WORK_QUIT (404)
19:10:15:WARNING:WU02:FS01:Server did not like results, dumping
19:10:15:WU02:FS01:Cleaning up

 

Clearly the results were not valid, was this a GPU slot?

[KenjiS9965]

5 hours ago, cbigfoot said:

dang so that 1080 ti must be a monster as you beat me badly.  look at my sig thats what im fighting with...

 

5 hours ago, tech.guru said:

i didnt buy for folding nvidia proprietary cuda technology handily beats amd in that regard.

 

5700 xt in games is great though, just wanted to give you that asterisk. for me its something i can donate to help out..

 

i wish could put the 8x tesla p100 at work to use. they would kill at folding.

I'm not sure its really that much of a monster, the 5700 XT is a fantastic card and in gaming is about equal if not a little faster depending on the title.. Seems as newer games come out the newer cards like the 5700 XT/2070 Super and 2080 Super are all starting to get larger and larger performance gaps over the 1080 Ti, I dont know if its an architectural thing that newer games properly utilize or if its simply far better driver optimizations for newer titles favoring the newer cards though...

[TrueMad]

Just now, GOTSpectrum said:

Clearly the results were not valid, was this a GPU slot?

Yes

[Ayren]

I just got a unit queued up for a GPU that aint even 50% done with its work. dafuq. I mean more work thats neat but muh bonus ;-;

[justpoet]

5 minutes ago, TrueMad said:

Yes

Try dropping the OC (even the factory one) on it a bit.  That usually prevents stuff like this from happening.  Compute use is typically harder on the cards than gaming is, so what was stable for gaming may not be completely so here.

[GOTSpectrum]

4 minutes ago, TrueMad said:

Yes

 

2 minutes ago, justpoet said:

Try dropping the OC (even the factory one) on it a bit.  That usually prevents stuff like this from happening.

Beat me to it. 

[GOTSpectrum]

@justpoet Core 22 is such a bitch with Ocin

[Ayren]

14 minutes ago, Ayren said:

I just got a unit queued up for a GPU that aint even 50% done with its work. dafuq. I mean more work thats neat but muh bonus ;-;

I guess ill try and put another gpu thats gonna finish soon and is still looking for a new unit on finish to try and let it drop its current search in the hopes that when i tell it to fold again itll switch to the other spare one

[Yujah]

1 minute ago, GOTSpectrum said:

@justpoet Core 22 is such a bitch with Ocin

Well... mis-rendering two pixels on CS:GO does seem to belong to another category of mishap than mis-computing two F@h Work Units...

[Ayren]

15 minutes ago, GOTSpectrum said:

@justpoet Core 22 is such a bitch with Ocin

Meanwhile:

and zero problems

[GOTSpectrum]

Just now, Ayren said:

Meanwhile:

and zero problems

on core 21 my 1080ti would run at 2075mhz, on core 22 it would so 2025. Plus the heat output is so much higher. It's nice as it means it's using more of the chip but still

[Ayren]

14 minutes ago, Ayren said:

and zero problems

And that card is even bios modded from the standard to the OC model (+30W power target)

[Ayren]

14 minutes ago, GOTSpectrum said:

on core 21 my 1080ti would run at 2075mhz, on core 22 it would so 2025. Plus the heat output is so much higher. It's nice as it means it's using more of the chip but still

Core 22 is pretty memory heavy... putting my memory OC to 550 was got bit of improvement. was untested though and i did it while it was running. guess i was lucky

[DirectorFlik]

This has been a fairly fun thing to join in and do.  I started off by giving my old 2nd gen i7 Falcon Northwest tower something to do.  Sure it's got a GTX580 and a 2600k processor, but it still wants to help

The computer itself has such a neat history that I can never justify getting rid of it.  Maybe one day I can give it some newer video cards.

 

I then started adding more random towers I've had sitting around.  There's a 1st gen i7 running in the garage that I tossed some old GTS 8800's into just to see if they could even do anything (I quickly learned NOPE), but at least the CPU can do something.  At this point there's also an i5 Optiplex using CPU, an i7 Dell machine of sorts doing CPU, and my main tower with a slower 8 core Xeon and my GTX 1080 humming along.  Nothing is overclocked, nothing is temperature monitored.  Just letting things roll along and do their thing.  I could probably keep adding more CPUs to the mix over the weekend here if I get too bored

[DelphinusVyse]

Question: People are mentioning getting another WU assigned before the current ones are finished. How can you see when that happens? Just curious.

[KenjiS9965]

3 minutes ago, DirectorFlik said:

This has been a fairly fun thing to join in and do.  I started off by giving my old 2nd gen i7 Falcon Northwest tower something to do.  Sure it's got a GTX580 and a 2600k processor, but it still wants to help

The computer itself has such a neat history that I can never justify getting rid of it.  Maybe one day I can give it some newer video cards.

 

I then started adding more random towers I've had sitting around.  There's a 1st gen i7 running in the garage that I tossed some old GTS 8800's into just to see if they could even do anything (I quickly learned NOPE), but at least the CPU can do something.  At this point there's also an i5 Optiplex using CPU, an i7 Dell machine of sorts doing CPU, and my main tower with a slower 8 core Xeon and my GTX 1080 humming along.  Nothing is overclocked, nothing is temperature monitored.  Just letting things roll along and do their thing.  I could probably keep adding more CPUs to the mix over the weekend here if I get too bored

Wonder if my old i7-3740QM and GTX680M Alienware M17x R4 could do something

[DirectorFlik]

2 minutes ago, DelphinusVyse said:

Question: People are mentioning getting another WU assigned before the current ones are finished. How can you see when that happens? Just curious.

I believe from what I've seen, you'll see an additional download for that folding slot before your existing WU for that slot is done.

[Ayren]

13 minutes ago, DelphinusVyse said:

Question: People are mentioning getting another WU assigned before the current ones are finished. How can you see when that happens? Just curious.

Spoiler

note how i have more work units than slots

its mainly caused by an option you can add to a slot called next-unit-percentage with valid values from 90 to 99 determining when to start looking for the next task after the current one

[DirectorFlik]

1 minute ago, KenjiS9965 said:

Wonder if my old i7-3740QM and GTX680M Alienware M17x R4 could do something

Worth a try!  It either works or it doesn't.  Always fun to give new live to old stuff.

[Yujah]

2 minutes ago, DelphinusVyse said:

How can you see when that happens?

In the "Advanced Control" interface. Either by directly monitoring the work queues, or in its Log tab (or in the Log directly).

[Guest]

2 minutes ago, Ayren said:

  Reveal hidden contents

note how i have more work units than slots

its mainly caused by an option you can add to a slot called next-unit-percentage with valid values from 90 to 99 determining when to start looking for the next task after the current one

Quick guys, shoot him, he's hogging all the resources!!! (=P)

[Ayren]

My extra system is ready to be configured

gonna be away for a bit :3

 

[TrueMad]

It's all about good cooling right ?    

So i have to try harder 

[Yujah]

3 minutes ago, KenjiS9965 said:

Wonder if my old i7-3740QM and GTX680M Alienware M17x R4 could do something

These specs: https://www.techpowerup.com/gpu-specs/geforce-gtx-680m.c344

 

do mention double-precision support ("FP64") for the GPU and basically any CPU will do (although it's far less efficient) so, yes, that would be useful.

 

 

[Ayren]

13 minutes ago, SpiritWrath said:

Quick guys, shoot him, he's hogging all the resources!!! (=P)

im about to hog a whole lot more