LTT Official Folding Month 2019!!!

[Stormseeker9]

6 hours ago, Kochiha Ichihara said:

The Planet Smasher lives up to its name.

You’re catching up fast!

 

Was 2.8 months 2 days ago

[Nathans nonsense]

how do i find my passkey again?

[shaz2sxy]

4 minutes ago, Nathans nonsense said:

how do i find my passkey again?

If you dont hae the email

 

goto the Passkey website 

 

https://apps.foldingathome.org/passkey/create

 

use the same info you did originally and you will get another email with the key

[Nathans nonsense]

13 minutes ago, shaz2sxy said:

If you dont hae the email

 

goto the Passkey website 

 

https://apps.foldingathome.org/passkey/create

 

use the same info you did originally and you will get another email with the key

oh i found it on the webcontrol website thing luckily

[sazrocks]

The production increase throughout September is barely even a bump now...

 

Also, our last update (12am) was only about 3.4m behind curecoin's. I think we will actually be able to catch them in production, at least for a little bit. I just want us to show up on their Threats chart, event if its only for a few days and has a time of 22 years, lol

[sazrocks]

On another note this latest update pushed me over 300,000,000 points total for the team!

[Stormseeker9]

Quote

07:10:12:WU00:FS02:0x21:Completed 20000 out of 2000000 steps (1%)
07:12:05:WU00:FS02:0x21:Completed 40000 out of 2000000 steps (2%)
07:13:58:WU00:FS02:0x21:Completed 60000 out of 2000000 steps (3%)
07:15:50:WU00:FS02:0x21:Completed 80000 out of 2000000 steps (4%)
07:17:43:WU00:FS02:0x21:Completed 100000 out of 2000000 steps (5%)
07:19:36:WU00:FS02:0x21:Completed 120000 out of 2000000 steps (6%)
07:21:38:WU00:FS02:0x21:Completed 140000 out of 2000000 steps (7%)
07:22:50:WU00:FS02:0x21:ERROR:exception: Error downloading array energyBuffer: clEnqueueReadBuffer (-5)
07:22:50:WU00:FS02:0x21:Saving result file logfile_01.txt
07:22:50:WU00:FS02:0x21:Saving result file log.txt
07:22:50:WU00:FS02:0x21:Folding@home Core Shutdown: BAD_WORK_UNIT
07:22:51:WARNING:WU00:FS02:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
07:22:51:WU00:FS02:Sending unit results: id:00 state:SEND error:FAULTY project:14225 run:170 clone:0 gen:34

What does this error mean about downloaded array energy buffer? encountered 2 of these in the last 2-4 days. It happens at the beginning of some WU's. Both happened within 5-8% completed.

[sazrocks]

2 minutes ago, Stormseeker9 said:

What does this error mean about downloaded array energy buffer? encountered 2 of these in the last 2-4 days. It happens at the beginning of some WU's. Both happened within 5-8% completed.

Generally it means instability, either from drivers or from overclocks. Do you have any overclocks?

[Stormseeker9]

11 minutes ago, sazrocks said:

Generally it means instability, either from drivers or from overclocks. Do you have any overclocks?

Yes I do. It was at first much worse, it was quiet for a 5 days then 1 of those popped up and 3 days later one too. Clocks / temps are all stable. I been tinkering with settings last few days and got it perfect seems daily ppd is increasing and some days f@h even reports 1.6m and in around 1.45m now 

[Sequence211]

16 minutes ago, Stormseeker9 said:

Yes I do. It was at first much worse, it was quiet for a 5 days then 1 of those popped up and 3 days later one too. Clocks / temps are all stable. I been tinkering with settings last few days and got it perfect seems daily ppd is increasing and some days f@h even reports 1.6m and in around 1.45m now 

It must be on the edge of stability then. try putting it one notch down.

[SnowyMus]

Starting a day late, but I've just got everything set up! It may not be Vega 64, or 2080 Ti level of performance, but I like to think that every little bit helps.

 

 

The other PC isn't going to be folding 24/7 as I need to use it during the day, and that PC is around 41% of the PPD shown here.

[Corrupt_Liberty]

Anybody watching this Cascade Lake X reveal?  Seriously contemplating an upgrade to i9-10980XE when it comes out.  Those prices are insane!  Thanks AMD!

[Sequence211]

7 minutes ago, Corrupt_Liberty said:

Anybody watching this Cascade Lake X reveal?  Seriously contemplating an upgrade to i9-10980XE when it comes out.  Those prices are insane!  Thanks AMD!

I am thinking about doing the same (or a lower SKU), I currently have a I7-6900k

[Corrupt_Liberty]

Just now, Struck said:

I am thinking about doing the same (or a lower SKU), I currently have a I7-6900k

Honestly I'd probably get the 12 core 10920X.  I'm currently running an i9-7900X so i'm hoping just a BIOS update and slap it in?  We'll have to see.  I'm really hoping they bring NVME RAID support to X299 without all the VROC BS.  I've got 2 NVME 1TB drives in there just waiting for a boost.

[Stormseeker9]

51 minutes ago, yaboistar said:

core overclock instability - had this on my 1080ti. it causes a driver reset and then it carries on as normal. drop your OC down

 

40 minutes ago, Struck said:

It must be on the edge of stability then. try putting it one notch down.

Dropped core clock from +165 to +160. Still runs at 2010mhz

Il'l check via remote in a few hours if it still occurs and drop to +155mhz

[Spotty]

RIP
I think I did a little too good of a job dusting my system yesterday.

GTX1080Ti was happily folding at 2037MHz at around 60-62°C all day. Was a little warm today, high 20s low 30s so I had the fans running 100%.
Came back to my computer after cooking dinner and GPU temp had dropped almost 10 degrees down to 53°C and the core clock was boosting up to 2062MHz. Wasn't long before it crashed with unstable clocks, went from 1.2M PPD estimate for the card down to 800K PPD estimate.
Will need to restart my PC to reset the card but I have a large file transfer (~600GB) backing up to my NAS and I can't restart until that's finished.

Damn you temperature dependent GPU Boost 2.0! 

[Stormseeker9]

Just now, yaboistar said:

that's not enough - rtx cards have a stepping of 15mhz - decrease it again until it steps down

 

are you running a memory overclock?

Ill remote in, wonder if it will make a big difference running 2010mhz v 1995mhz in completion of WU's.

Memory is +500 so should run at 7500mhz but only running 7300mhz at the moment ( which I cannot explain)

[shaz2sxy]

3 minutes ago, Stormseeker9 said:

Ill remote in, wonder if it will make a big difference running 2010mhz v 1995mhz in completion of WU's.

Memory is +500 so should run at 7500mhz but only running 7300mhz at the moment ( which I cannot explain)

Just keep memory at stock. I am on my 2070 super

 

+100 on core

120% power target

[Stormseeker9]

Just now, yaboistar said:

because of p2 state.

1995 will be fine, your chip seems to have lost the sillicon lottery a bit.

Whats P2 state?

THe memory going down to 7300 mhz from 7500mhz was after I cleaned my PC a few weeks back. Since then my 2070 also idles at 59c... 

Wonder if I damaged it somehow? Got one of those PC leafblowers 

[Stormseeker9]

3 minutes ago, yaboistar said:

idling at 59c is normal for zero-fan cards.

 

p2 state is an internal gpu power state which determines that the core frequency is allowed to boost as normal but memory frequency will be capped. it's enabled byy default for compute tasks to aid stability in those compute applications

so just remote accessed my PC at home Dropped memory OC to 0 and its only running at 6801Mhz instead of the stock 7000mhz.. Surely something is off here?

 

 

Before my GPU was idling in the mid to high 30's... hence 59c seems odd to me.  

 

 

 

5 minutes ago, shaz2sxy said:

Just keep memory at stock. I am on my 2070 super

 

+100 on core

120% power target

Dropped one step from previous OC so +150.

Memory to 0, but running onder stock speed now..

[Nathans nonsense]

9 hours ago, Nathans nonsense said:

someone tell my computer it has to get me over 700k points a day like it used to or i'm not buying it a new motherboard for Christmas

  yo wth it worked lol

[Stormseeker9]

1 minute ago, yaboistar said:

i'd say that's about normal for p2 state - it does have a slight reduction in memory speed over stock

 

are your load temps any different from normal? is the fan speed needed to maintain those load temps higher than it normally was?

 

depending on what the card is doing, my 1080ti can idle anywhere from 35-60c based on whether i've got things like chrome or origin open which are utilising the GPU, even in the background

No, under load the temps seem the same. Run fans on auto normally. Highest temp under load i've seen is I think 71C and that was when it was hotter over here in europe. Usually under load its mid - high 60's. (1440p res)

 

Got fans ramped up now to keep the card a bit cooler in the mid 60's instead of pushing 70-72c.

[Stormseeker9]

4 minutes ago, yaboistar said:

that might be why you're crashing - WU finishes, temperature drops really low because fans are set to a constant high value, GPU boost slams the core speed way up high because it can see it has the thermal headroom and then it crashes when it comes back onto load because compute tasks are super sensitive

Got fans at 65C.

I do have seen those drops like 10c delta. and then directly jumping back up to under load temps. fans are running 65%. I noticed on auto is was usually around the 53-55% fan speed.

 

For now I dropped Core to +150mhz took away the memory clock all together. PL at 114% cV +66.

Hope my PPD doesnt drop too much / time for completion of WU's

[Stormseeker9]

4 minutes ago, yaboistar said:

it won't drop much, no. power limit and voltage sliders maxed with a very conservative core bump is how i run all my cards.

Got it setup like this now.

 

PPD did drop from 16xxxxx to what its displaying now. Hope this way no more errors, otherwise just gotta live with 'm 

[Stormseeker9]

6 minutes ago, yaboistar said:

slow WU's are better than errored WU's, as if WU's return faulty they're essentially useless and a waste of time for you and the project.

 

also bear in mind that PPD fluctuates quite a lot based on a whole bunch of factors including system load. a WU for me can say anything from 3.2 to 3.6mppd and will fluctuate around those numbers while it's completing

Yeah thankfully these last 2 errors occured about 5-8% in the WU and not halfway through.