LTT Folding Team's Emergency Response to Covid-19

[LFreeman]

12 minutes ago, Kilrah said:

I can validate, 400K-ish PPD is about what I get with my 970 (well got, pulled it out earlier since it was causing issues and not really worth it given my other GPUs)

when did you get Silver?

[Gorgon]

16 minutes ago, Cabanaj said:

Thanks alot @Gorgon for that info on PPD! I did try to check about that, and ppd/watt, could not fint it anywhere! That will simplify things. I might just give the card (GTX 970) to someone That can do something with it. I try to waste as little as possible.

As for the OEM part, I should have make the precision: When I was talking about buying a fans module, I was refering to Newegg or ebay. Only ones I find are around 50$ and on Newegg, 30$ + 59.95$ shipping... it make it 90$.

You can find the approximate PPD in that google sheet I linked and power usage can be found in the TechPowerUp GPU database. Currently Turing is the champion for PPD/W. I'd like to see more competition from AMD but it appears with Navi their drivers and/or OpenCL implementation is lacking.

[Kilrah]

6 minutes ago, LFreeman said:

when did you get Silver?

3 days ago

 

Got another 9M since

 

[Gorgon]

Those RX580s have been working hard

[Egg-Roll]

So after playing around with my Ryzen CPU for a little bit (and crashing my computer if I load Ryzen Master ) I have decided to just turn off all boost and over clocking, having a CPU at 3.9Ghz 24/7/5+years at 75C tops under load is safer than having a CPU at 90+C for possibly as long with almost no benefit... Also doesn't change my points or time to complete much if any...

[LFreeman]

4 minutes ago, Kilrah said:

3 days ago

 

Got another 9M since

 

Yeah I wonder why they haven't gotten to my request yet.

[Unilevers]

12 minutes ago, Gorgon said:

Those RX580s have been working hard

haha, congrats and yes those 580's must be working very hard!

[Gorgon]

1 hour ago, calberto said:

I have 2 client names because the Windows client will not allow spaces. I Imagine it's too late for corrections?

Likely but these two will be able to tell you

@GOTSpectrum @leadeater

[Gorgon]

Looks like Team EVGA will be dropping out of 2nd place overall in the coming week.

[jctappel67]

2 hours ago, Favebook said:

 

Until you realize google has banned your card from making more accounts...

I had up to three accounts on same card when I was doing cloud F@H last event, but it was dangerous play as Google started banning them one by one even after they were previously accepted.

Also, I burned all my cards and family's cards last event, not sure if I can use them again for free credit (I would assume not...)

That's why I have another card lined up for the new ones.

 

I'm not really planning to keep cranking up for future events anyway. This is a fun distraction, but I'm still in college. I would love to pour money into dedicated folding rigs, and loads of 2080TIs, but I'm having a hard time convincing myself that I should upgrade from a 2060 to a 2070 or 2070S. 

 

Going all out for this event, and then I'll just turn it on when I feel like it moving forward.

[Bitter]

53 minutes ago, ObeliskAG said:

I have a 390X (275 watts) and if your workload requires double-precision computer, then like the Vega 64 (295 watts), it's a monster. However, Folding@Home is only using single precision, so these cards don't particularly stand out and are not particularly efficient. Assuming the Vega 64 is sustaining 1,546 MHz boost core clock, then that's a theoretical 12,665 GFLOPS for 42.93 GFLOPS per watt. By comparison a 2080 Ti can theoretically produce 64.78 GLOPS/watt - that's almost 51% more efficient than a Vega 64! (Boost core frequency in GHz x number of shaders x 2 = Theoretical single-precision GFLOPS)

 

Makes sense long term, up front cost of a Vega 56 is pretty low though which sets that cross over point farther out into the future depending on your local electric rates and depending on actual wattage being consumed. My Vega 56 is using about 180W and my RX 580 is using about 125W, as reported by HWinfo, while folding. So no where near those numbers of power use but assuming that since I'm bench marking around what stock Vega 64's benchmark at that I'm doing within a few percent the same gflops as a 64 does then I'm closer to the 2080ti than I am to the Vega numbers, right? I've got about 100mhz OC from factory and an undervolt I've applied, plus when folding it never reaches that full power draw since it's not using all of the GPU core for work like it would be doing for heavy 3D work. Even running Luxmark it doesn't get but into the 220-230 watt range and scores around with a 1080ti and Vega 64 score. Not really sure how Luxmark scores translate to folding PPD or other metrics, but it's a quick easy way to compare the compute power of different cards at a glance since it's cross platform.

[walderston]

3 hours ago, Full_Metal_Jacob said:

Whats the largest amount of credits you guys have received from a WU?

 

My GPU is working on a 98K one right now.

Project: 14411

WU00:FS00:Final credit estimate, 445264.00 points

 

[Gorgon]

2 minutes ago, walderston said:

Project: 14411

WU00:FS00:Final credit estimate, 445264.00 points

 

Is that one of your RX580s?

[cbigfoot]

3 minutes ago, Gorgon said:

Is that one of your RX580s?

no thats his r7-350x

[walderston]

7 minutes ago, Gorgon said:

Is that one of your RX580s?

Its actually a ATI Radeon B530  Box full of them. 

 

In all seriousness, its a 2080 TI with factory OC from Gigabyte

[Egg-Roll]

1 minute ago, Gorgon said:

Looks like Team EVGA will be dropping out of 2nd place overall in the coming week.

Didn't they change their policies a year or so ago making it not really worth while doing because the credit expired in a year or something like that? Idk which one would be the better deal Curecoin or EVGA and I think that is what has caused them to loose numbers. Not to mention they are only giving out 120 "bucks" per 30 mill points ($10 a month), so if someone wanted a new GPU for a low price with my current setup one could get about $700 from Cure in 1 year profit vs $60 because I won't ever hit 30 million.

[GOTSpectrum]

2 hours ago, LFreeman said:

what platform?

Check my profile! 

 

[GOTSpectrum]

26 minutes ago, Gorgon said:

Likely but these two will be able to tell you

@GOTSpectrum @leadeater

The correction deadline was around noon on the first day... Like a week and a half ago. @calberto

[GOTSpectrum]

23 minutes ago, Gorgon said:

Looks like Team EVGA will be dropping out of 2nd place overall in the coming week.

F

 

But seriously its about time! 

[HaRdLy007]

Are us peons going to see the F@H server install and setup, or is it a FP only thing?

[Gorgon]

4 minutes ago, Egg-Roll said:

Didn't they change their policies a year or so ago making it not really worth while doing because the credit expired in a year or something like that? Idk which one would be the better deal Curecoin or EVGA and I think that is what has caused them to loose numbers. Not to mention they are only giving out 120 "bucks" per 30 mill points ($10 a month), so if someone wanted a new GPU for a low price with my current setup one could get about $700 from Cure in 1 year profit vs $60 because I won't ever hit 30 million.

Doesn't really matter as their shipping costs are insane for anywhere other than the continental US. For Canada, when I last checked, it would be $50US for shipping as they insist on using one of the more expensive couriers.

[Gorgon]

1 minute ago, HaRdLy007 said:

Are us peons going to see the F@H server install and setup, or is it a FP only thing?

Part 1 was posted on FP 5 days ago so I believe it will be on uTube in a couple of days. No sign of Part 2 yet though the server seems to be in full production now and has been upgraded to 9.6.1 tho I see some servers at 9.6.3

[ObeliskAG]

29 minutes ago, Bitter said:

Makes sense long term, up front cost of a Vega 56 is pretty low though which sets that cross over point farther out into the future depending on your local electric rates and depending on actual wattage being consumed. My Vega 56 is using about 180W and my RX 580 is using about 125W, as reported by HWinfo, while folding. So no where near those numbers of power use but assuming that since I'm bench marking around what stock Vega 64's benchmark at that I'm doing within a few percent the same gflops as a 64 does then I'm closer to the 2080ti than I am to the Vega numbers, right? I've got about 100mhz OC from factory and an undervolt I've applied, plus when folding it never reaches that full power draw since it's not using all of the GPU core for work like it would be doing for heavy 3D work. Even running Luxmark it doesn't get but into the 220-230 watt range and scores around with a 1080ti and Vega 64 score. Not really sure how Luxmark scores translate to folding PPD or other metrics, but it's a quick easy way to compare the compute power of different cards at a glance since it's cross platform.

What's your sustained core clock on the Vega56?

[LFreeman]

@GOTSpectrum I should have thought to look at that

[LFreeman]

Anyone seen this error before? It keeps happening.

 

FahCore returned: INTERRUPTED (102 = 0x66)