Folding Community Board

[TheLANguy]

58 minutes ago, rkv_2401 said:

They're doing it for rewards, or at least what they perceive as a reward. Doesn't look like it's going too well, though. https://www.solinscription.com/care-market/

 

This is the only way to offload the CARE tokens, but they would need a buyer and the buyer would need some utility for the tokens, which ATM doesn't seem to exist.

Seems like a temporary boom.

[GOTSpectrum]

28 minutes ago, TheLANguy said:

Seems like a temporary boom.

Most probably

[madprofessor207]

On TechLinked they talked about Zluda, which allows CUDA to run on Intel and AMD GPU's. It was abandoned by AMD, but it is opensource. Could it be used to make AMD Cards better at folding CUDA WU's on AMD cards? Or for Boinc?

 

[unijab]

Im getting close to the next badge (bronze) ...

 

[unijab]

2 hours ago, madprofessor207 said:

Could it be used to make AMD Cards better at folding CUDA WU's on AMD cards? Or for Boinc?

I'd like to try it out...

 

will probably wait until i get my bronze badge first

[BondiBlue]

Man, I've been busy lately. I finally had a chance to get F@h working on my 2070 Super + 1080 machine, and things are going well once again. F@h had stopped working a few months ago for no apparent reason, so this machine had been out of action for quite some time. It's now back up and running, and I'm planning to get my dual Quadro P4000 machine back up soon. 

[RollinLower]

15 hours ago, BondiBlue said:

Man, I've been busy lately. I finally had a chance to get F@h working on my 2070 Super + 1080 machine, and things are going well once again. F@h had stopped working a few months ago for no apparent reason, so this machine had been out of action for quite some time. It's now back up and running, and I'm planning to get my dual Quadro P4000 machine back up soon. 

welcome back, i can't imagine how chilly your place has been these months without F@H to keep you warm!

[unijab]

either tonight or tomorrow...

 

[Gorgon]

This is a new one. I have 6 slots stuck Sending but the twist is it doesn't appear to be a network issue but with the server:

core:0x22 unit:0x360300000c0200000000000025490000
15:18:19:WU01:FS00:Uploading 34.54MiB to 131.239.113.97
15:18:19:WU01:FS00:Connecting to 131.239.113.97:8080
15:18:25:WU01:FS00:Upload 19.90%
15:18:31:WU01:FS00:Upload 41.61%
15:18:37:WU01:FS00:Upload 63.50%
15:18:43:WU01:FS00:Upload 85.40%
15:18:47:WU01:FS00:Upload complete
15:18:47:WU01:FS00:Server responded PLEASE_WAIT (464)
15:18:47:WARNING:WU01:FS00:Failed to send results, will try again later

My other slots, connected to other Servers, all uploaded successfully it just appears to be an issue with 131.239.113.97. The Server Stats show the unit has one error - Out of Disk Space. Whoops!

 

There is already a post on the Support Site about the issue.

 

I guess we'll just have to watch the QRB dwindle until they get it fixed.

[DreamCat04]

I decided to use my gaming laptop as a folding machine as I have a gaming desktop already that I mainly use and I actually only use that laptop once a week, so I thought why not make it into a folding "server". I'm only folding on the GPU so that I don't stress it too hard. If I limit the CPu to 3.6 GHz by setting the Speed Shift EPP to 130 in ThrottleStop, the GPU (GTX 1660Ti 110W TGP) can draw its full 95W while running at around 56°C above ambient temp (with a current room temp of 22°C that results in a GPU temp of 78°C and the hotspot is at around 87°C). My estimated PPD is at around 280k like this. My 12700K in the desktop does around 85k PPD at 200W of power draw with my OC of 5GHz all-core

[rkv_2401]

21 hours ago, DreamCat04 said:

I decided to use my gaming laptop as a folding machine as I have a gaming desktop already that I mainly use and I actually only use that laptop once a week, so I thought why not make it into a folding "server". I'm only folding on the GPU so that I don't stress it too hard. If I limit the CPu to 3.6 GHz by setting the Speed Shift EPP to 130 in ThrottleStop, the GPU (GTX 1660Ti 110W TGP) can draw its full 95W while running at around 56°C above ambient temp (with a current room temp of 22°C that results in a GPU temp of 78°C and the hotspot is at around 87°C). My estimated PPD is at around 280k like this. My 12700K in the desktop does around 85k PPD at 200W of power draw with my OC of 5GHz all-core

i) Disable turbo boost instead of playing with speed shift when you're folding, it's easier and provides a lot more thermal headroom

 

ii) A 1660Ti mobile should get 1-1.4M PPD, why do you only get 280k?

[DreamCat04]

On 2/18/2024 at 12:52 PM, rkv_2401 said:

i) Disable turbo boost instead of playing with speed shift when you're folding, it's easier and provides a lot more thermal headroom

 

ii) A 1660Ti mobile should get 1-1.4M PPD, why do you only get 280k?

Perhaps it was because I closed the lid of the laptop, which makes the CPU clock down to 800Mhz even though I set it to do nothing when the lid is closed. That results in the GPU dropping down to around 65W from 95W, might have to re-test it with the lid open (I did have the lid open for one WU but didn't re-check the PPD. Disabling Turbo boost actually limits my CPU further, to just 2.6GHz, so with SpeedShift at 130, I have around a GHz more while the power consumption is lowered so far that the GPU draws its full 95W. When the CPU is drawing 30W for example, the GPU only does around 80-85W. I need the CPU to stay below 20W to get the full 95W on the GPU

Edited February 20 by DreamCat04

[rkv_2401]

I'm glad to hear that the SpeedShift setting works for you, but if you're not doing anything else heavy duty on the laptop when you're folding, you can save another 10W and still get the same PPD. I'm not sure how important power efficiency is to you, though. My set ups are all optimized towards efficiency.

Also, I get 1-1.4M PPD with my GPU reporting 60-70W. My laptop has an i7-10750H with a 1660Ti (90W), almost the same as yours but I've undervolted both the CPU and GPU. I believe if your GPU is boosting fully it should be around 1.8M PPD. Look into undervolting if you're interested in efficiency, it'll help your laptop stay cool and cut the bills by a bit too!

[RollinLower]

i just checked the production on my main machine, it doesn't seem to have recovered PPD-wise after the first issue with the collection server running out of space.

i don't see any failed uploads anymore, so i'm honestly not sure what's going on now. all the logs look clean with downloads, work being completed and uploads all working as intended. Just my PPD is down almost 50%

 

any ideas what could be causing this?

 

EDIT: also doublechecked my passkey, that checks out.

[Gorgon]

1 hour ago, RollinLower said:

i just checked the production on my main machine, it doesn't seem to have recovered PPD-wise after the first issue with the collection server running out of space.

i don't see any failed uploads anymore, so i'm honestly not sure what's going on now. all the logs look clean with downloads, work being completed and uploads all working as intended. Just my PPD is down almost 50%

 

any ideas what could be causing this?

 

EDIT: also doublechecked my passkey, that checks out.

Likely just a backlog of work units that need to be processed. Hopefully you'll see a big update shortly.

 

After the 4 slots I had that were stuck trying to upload It's been smooth sailing here and I haven't noticed any major decrease in production being reported,

[RollinLower]

12 hours ago, Gorgon said:

Likely just a backlog of work units that need to be processed. Hopefully you'll see a big update shortly.

 

After the 4 slots I had that were stuck trying to upload It's been smooth sailing here and I haven't noticed any major decrease in production being reported,

it's been a while though, lets hope PPD picks back up soon.  

[RollinLower]

On 2/21/2024 at 10:44 PM, Gorgon said:

Likely just a backlog of work units that need to be processed. Hopefully you'll see a big update shortly.

 

After the 4 slots I had that were stuck trying to upload It's been smooth sailing here and I haven't noticed any major decrease in production being reported,

honestly not sure if this is still backlog. is there anything i can do to see if F@H is properly accepting my passkey and granting QRB? i can't really find anything in the logs.

right after that capacity problem at F@H my PPD has halved almost exactly, and it's consistent in the new halved PPD too!
 

[Gorgon]

2 hours ago, RollinLower said:

honestly not sure if this is still backlog. is there anything i can do to see if F@H is properly accepting my passkey and granting QRB? i can't really find anything in the logs.

right after that capacity problem at F@H my PPD has halved almost exactly, and it's consistent in the new halved PPD too!
 

Here are your current "CPUs" in use (actually CPUs and GPUs) which show 2 CPUs and 3 GPUs currently folding, all of which are receiving the QRB. There is a AMD GPU which appears to be MIA since Feb 17th and a x86 GPU missing since the 18th, but that may just be a a slot being re-configured.

 

I'd check at this point to see if you have a bunch of failed WUs. HfM is useful for this, otherwise you'll have to search the logs for "CORE_RESTART" "NaN" "exception occurred" The Log Viewer in the Advanced Client can be useful for this by checking "Warnings & Errors"

[rkv_2401]

According to LARsystems' database, the RTX 4070 Super gets between 10-13M PPD, but the 3070 gets just 5M PPD? Has there really been a (more than) doubling of PPD in this generation?

[Gorgon]

6 hours ago, rkv_2401 said:

According to LARsystems' database, the RTX 4070 Super gets between 10-13M PPD, but the 3070 gets just 5M PPD? Has there really been a (more than) doubling of PPD in this generation?

The 3070 has 5888 Shaders while the 4070 Super has 7168. So not exactly an apples to apples comparison. You'll want to compare it to the 4070 as that has the same number of shaders. So ~7.9MPPD versus 4.7MPPD.

 

Yes, 4000-series GPUs are astounding for both their efficiency and price. Though the base clocks have increased from ~1600 to ~2000MHz the power consumption has also gone way down so not only do they perform well but they can run a lot cooler than Ampere.

[RollinLower]

15 hours ago, Gorgon said:

Here are your current "CPUs" in use (actually CPUs and GPUs) which show 2 CPUs and 3 GPUs currently folding, all of which are receiving the QRB. There is a AMD GPU which appears to be MIA since Feb 17th and a x86 GPU missing since the 18th, but that may just be a a slot being re-configured.

 

I'd check at this point to see if you have a bunch of failed WUs. HfM is useful for this, otherwise you'll have to search the logs for "CORE_RESTART" "NaN" "exception occurred" The Log Viewer in the Advanced Client can be useful for this by checking "Warnings & Errors"

no failed WU's anywhere. just WU's finishing and work being completed as usual. i'm genuinely at a loss. i guess i'll just try to reinstall FAH at this point. no clue if it could help but it won't hurt to try.

[RollinLower]

1 hour ago, RollinLower said:

no failed WU's anywhere. just WU's finishing and work being completed as usual. i'm genuinely at a loss. i guess i'll just try to reinstall FAH at this point. no clue if it could help but it won't hurt to try.

i upgraded to the beta V8 client while i was at it, so far PPD seems to be pretty simmillar but lets see what happens after it crunched a couple WU's.

 

sidenote, the difference between V7 and V8 in functionality is pretty astronomical! i love the way machine management has been moved off-node so if you have multiple boxes running the process and need for 3rd party apps has been removed completely. i wonder how this will work in the long run.

[RollinLower]

3 hours ago, RollinLower said:

i upgraded to the beta V8 client while i was at it, so far PPD seems to be pretty simmillar but lets see what happens after it crunched a couple WU's.

sidenote, the difference between V7 and V8 in functionality is pretty astronomical! i love the way machine management has been moved off-node so if you have multiple boxes running the process and need for 3rd party apps has been removed completely. i wonder how this will work in the long run.

it completed about 1 to 3 WU's on each slot now, and still PPD remains down there. What i also notice is that GPU utilization remains low even after upgrading all my drivers.

 

seems like there is just still something missing i'm not seeing. this setup should get me about 30M to 40M PPD but it remains stuk on just 15M.

[Gorgon]

2 hours ago, RollinLower said:

it completed about 1 to 3 WU's on each slot now, and still PPD remains down there. What i also notice is that GPU utilization remains low even after upgrading all my drivers.

 

seems like there is just still something missing i'm not seeing. this setup should get me about 30M to 40M PPD but it remains stuk on just 15M.

Hmm. Those Wattage and Utilization numbers look low for the 4070tis. Could it be a driver issue? Are you Power or Clock Limiting? (I doubt it as Persistence Mode is off.) Here's what two of my 4070tis look like (p18219 & p1812226), both at ~12.9MPPD Clock-limited to 2205MHz:

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.147.05   Driver Version: 525.147.05   CUDA Version: 12.0     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA GeForce ...  On   | 00000000:0A:00.0 Off |                  N/A |
| 66%   76C    P2   150W / 285W |    502MiB / 12282MiB |     89%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
|   1  NVIDIA GeForce ...  On   | 00000000:0B:00.0 Off |                  N/A |
| 32%   59C    P2   138W / 160W |    496MiB / 12282MiB |     98%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+

[RollinLower]

35 minutes ago, Gorgon said:

Hmm. Those Wattage and Utilization numbers look low for the 4070tis. Could it be a driver issue? Are you Power or Clock Limiting? (I doubt it as Persistence Mode is off.) Here's what two of my 4070tis look like (p18219 & p1812226), both at ~12.9MPPD Clock-limited to 2205MHz:

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.147.05   Driver Version: 525.147.05   CUDA Version: 12.0     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA GeForce ...  On   | 00000000:0A:00.0 Off |                  N/A |
| 66%   76C    P2   150W / 285W |    502MiB / 12282MiB |     89%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
|   1  NVIDIA GeForce ...  On   | 00000000:0B:00.0 Off |                  N/A |
| 32%   59C    P2   138W / 160W |    496MiB / 12282MiB |     98%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+

no limits set, all settings are stock.

i tried limiting clocks to 2205 aswell but that yielded no PPD, just a bit lower in temps.

 

i did purge the old driver and clean install a the latest one, but i see you are on 525.147, which is quite a bit behind mine at 545. i'll see if downgrading works.