LTT Official Folding Month 2019!!!

[sanya567xxx]

8 hours ago, yaboistar said:

ran at 105c

vega64s can easily go to almost 120 TJ...

[sanya567xxx]

57 minutes ago, Struck said:

running 14180 WU

I have 14180 and 14191 almost exclusively on all my 3 1080tis

rarely getting like 14249 and 14250

Also apparently my worse-cooler-design card is stable at 2076MHz 1.081v.. The better cooler card isn't stable 2063MHz at even 1.150v

[GrockleTD]

9 minutes ago, sanya567xxx said:

I have 14180 and 14191 almost exclusively on all my 3 1080tis

rarely getting like 14249 and 14250

Also apparently my worse-cooler-design card is stable at 2076MHz 1.081v.. The better cooler card isn't stable 2063MHz at even 1.150v

feels lottery man

[Erudito]

Anyone else getting an abundant amount of WUs from Temple University for their processors to work on? 

[Manmeyco]

5 hours ago, Ryan_Vickers said:

There were already concerns about having event badges causing too much clutter, the last thing we need is this! XD

You can never have too many badges

[palespartan]

4 hours ago, GOTSpectrum said:

Yeah, I agree, from me a PM some time

I'm down to help how I can.

3 hours ago, TVwazhere said:

At this rate I'm gonna lose 20th....

I keep getting 14180 WU's and I've been pausing folding on the weekend to play games with my friends cause I'm too lazy to reconfigure the Main rig to use the iGPU on my 4790k, and trying to fold AND play Risk of Rain 2 caused my system to lock up. 

Yeah I have a feeling they're pushing this project hard right now. On my 2070 on Linux its only using 85% gpu and dropped from 1.6 to 1.3m ppd. Basically the full power of one of my 970's

[kropki]

tl;dr; too many pages keep folding.

[GOTSpectrum]

1 minute ago, kropki said:

tl;dr; too many pages keep folding.

All the pages!

 

Welcome to the thread, fold on brother!

[shaz2sxy]

Thats another day done!! i'm off to bed.... hopefully i'm still in the top 30

 

fingers crossed!!!

[Erudito]

So long as I stay in the top 300 I am good

[leadeater]

4 hours ago, TVwazhere said:

At this rate I'm gonna lose 20th....

Firewall block is being removed today, I'mma coming for ya 

[Gorgon]

1 hour ago, leadeater said:

Firewall block is being removed today, I'mma coming for ya 

Get the lead out

[Egg-Roll]

So I've just OC'd my 1070 and idk how much I can OC it before it crashes the computer/drivers... If anyone has a idea it would be greatly appreciated, just remember it is a non RMA/Thermal Pad Mod GTX 1070 FTW Model.

 

Tho I have my doubts this set up will burn down where I live it's sitting nice at 70C or less.

 

[sanya567xxx]

13 minutes ago, Egg-Roll said:

burn down

TLDR: it's safe on your card, worst problem is the need to reinstall driver in very bad cases. haven't happened once to me in 3 years tho. play around with numbers, set power and temp limits to max and just check which core clock is stable, then take it back a step. (cause the turboboost will mess with clocks anyway and sometimes try to run it higher than you set it to)

 

recent cards are so restrictive it's near impossible to damage them without doing hardware or bios modding.

so basically test to which frequency you're stable, then take it down a notch and run like that easy.

with mem - people say it doesn't impact on fah, but I run +250 on my 1080tis anyway, it seems to work better than +400 did (for some reason I don't get) and it didn't impact stability in my cases.

also if you go to afterburner settings - monitor, you can scroll down and see "gpu 1 - power limit", "gpu 1 - thermal limit", "gpu 1 - no load limit" and "gpu 1 - voltage limit" graphs which may show you what's limiting your performance. basically, if it's no load - unlucky WU, if it's voltage and you're already at max voltage (1.093 for stock bios pascal cards) - well Fs in chat, can't do nothing, if you're at power or thermal limit though you can adjust the power/termal target sliders (which i personally have maxed out at all times)

core voltage mV slider doesn't do s*** on Pascal or Turing cards usually (as Buildzoid aka Actually Hardcore OverClocking on YT says, it does, but under temperatures that are out of reach for most of the users - e.g. below 30C, that's not room, that's GPU die temp needed), but I've noticed it can worsen your stability in some occasions apparently. I leave mine 100% on the card that won't go higher than 1.093V as it's stock bios, and I know it's good at both 1.081 it runs usually and at 1.093), but I have it at 0% on the card with extreme bios cause there it somehow brought instability when not at 0% even though all the monitoring tools reported the same values as they did before anyway, but on that card I just control the voltage manually by adjusting the voltage-frequency curve (ctrl+f in afterburner) instead of using the core clock slider)

[LaserLight]

I'm surprised I'm in the top 100.  Go Radeon VII go!  Getting roughly 1.7- 1.8m PPD on it.  

[justpoet]

1 hour ago, leadeater said:

Firewall block is being removed today, I'mma coming for ya 

BOOM!  Here comes the BOOM!

[Egg-Roll]

13 minutes ago, sanya567xxx said:

TLDR: it's safe on your card, worst problem is the need to reinstall driver in very bad cases. haven't happened once to me in 3 years tho. play around with numbers, set power and temp limits to max and just check which core clock is stable, then take it back a step. (cause the turboboost will mess with clocks anyway and sometimes try to run it higher than you set it to)

 

recent cards are so restrictive it's near impossible to damage them without doing hardware or bios modding.

so basically test to which frequency you're stable, then take it down a notch and run like that easy.

with mem - people say it doesn't impact on fah, but I run +250 on my 1080tis anyway, it seems to work better than +400 did (for some reason I don't get) and it didn't impact stability in my cases.

also if you go to afterburner settings - monitor, you can scroll down and see "gpu 1 - power limit", "gpu 1 - thermal limit", "gpu 1 - no load limit" and "gpu 1 - voltage limit" graphs which may show you what's limiting your performance. basically, if it's no load - unlucky WU, if it's voltage and you're already at max voltage (1.093 for stock bios pascal cards) - well Fs in chat, can't do nothing, if you're at power or thermal limit though you can adjust the power/termal target sliders (which i personally have maxed out at all times)

core voltage mV slider doesn't do s*** on Pascal or Turing cards usually (as Buildzoid aka Actually Hardcore OverClocking on YT says, it does, but under temperatures that are out of reach for most of the users - e.g. below 30C, that's not room, that's GPU die temp needed), but I've noticed it can worsen your stability in some occasions apparently. I leave mine 100% on the card that won't go higher than 1.093V as it's stock bios, and I know it's good at both 1.081 it runs usually and at 1.093), but I have it at 0% on the card with extreme bios cause there it somehow brought instability when not at 0% even though all the monitoring tools reported the same values as they did before anyway, but on that card I just control the voltage manually by adjusting the voltage-frequency curve (ctrl+f in afterburner) instead of using the core clock slider)

It's why I'm not touching any power settings and just the Mhz, also mV is grayed out for my card in AB

 

As for the impossible...

https://www.guru3d.com/news-story/evga-ftw-1080-and-1070-have-overheating-issues.html

 

TBH I think the only reason why mine hasn't gone miniature nuclear meltdown is because there is less than 1-1.5cm distance from my massive (overkill Cooler Master Hyper 612 V2) cpu fan conveniently parked where the heating issue lies  Also have one bottom fan blowing up, a front fan blowing in, and a back fan right after the CPU cooler blowing out. I'ma going to see if I can add another fan blowing directly onto it  I'm in no short supply of 2000 rpm fans thanks to my new (to me) server

[MrSquishie]

Just wish to share a little news that Folding@Home has put up-
foldingathome.org - BONUS RECREDIT

[sanya567xxx]

36 minutes ago, Egg-Roll said:

As for the impossible...

 

lottery. if his power supply is absolute garbage, he removed half the board from the card (because there's a lot of circuitry that prevent over current, over temperature, over voltage, over power (general) for all the parts there, for core, for memory, for the power stages, for all kind of things, then that could happen. t h e o r e t i c a l l y, once in a life time

my card had it's thermal interface dry out, contact not being full, one of the fans not working, and I was blasting it with 100% load on OC settings. I was trying to figure out, what was resulting bad in bad performance, and noticed it was 95C and throttling to 1400 MHz instead of over 2000. It was still working fine for long time, so I fixed the fan, fixed the paste, and it's all good.

So those occasions are examples of being unlucky with some hardware inside the card. under normal circumstances (and you've already tested them via folding extensively) that would never happen.

Seriously though, your card is doing under 180W, probably even less than 150. It's hard to start a fire with that even.

 

 

 

[sanya567xxx]

3 hours ago, GOTSpectrum said:

All the pages!

 

btw is day 7 blog entry coming?

[Egg-Roll]

13 minutes ago, sanya567xxx said:

Seriously though, your card is doing under 180W, probably even less than 150. It's hard to start a fire with that even.

Very true, I've also done crypto on this thing too knowing full well and kept a eye on it as well.

 

However never can have too much fans, the question is should this 210mm fan push or pull  I'm leaning pull because it provides a big vent and sits in front of the card (assuming it fits and doesn't have to sit outside)

[sanya567xxx]

5 minutes ago, Egg-Roll said:

push or pull

just test. Or imagine how the air would move (or draw that) and see how it'd change the air pathway. Basically you don't want it to be looking for exit or having to change direction of movement. And you don't want to fans pointing at each other, like with traffic on the roads.

I'd also probably not add more fans until it's needed but lock them to 100%. 

Also, anti-vibrational resin things to hold fans instead of screws are SO GOOD. The fans stop resonating the metal case and sound quiter, especially at lower speeds, and also my table was previously vibrating because the pc was on it, I took some foam the PSU (or some other part, I don't even recall correctly) was shipped with, cut out pieces a little big then the case feet, and placed the pc on those pieces on top of the table. A lot less inconvenience now.

[TensorVortex]

all my gpu on 14180, I will be sleeping through this pain

[Devilish Husky]

just wanted to chime in and say good luck to everyone and I'm folding for Alzheimer's. wish I had a 3950x right now ;_;

[Metallus97]

7 hours ago, Yabdat said:

Ack, lost 10th place..

 

Time to dig into my literal pile of old hardware...

 

<One hour later...>

 

 

Hmm, didnt think I could get both running in the same system..

But HEY!, more points. 

do you have drivers installed for both?