firedfly

I hit top 100 folders according to EOC stats with my contributions to EHW.  I've also folded over 16 billion points for the team!

https://folding.extremeoverclocking.com/individual_overtake.php?s=&u=1248729&p=1

My total points for all teams puts me at 87th and 19 billion points in the fah stats!

https://stats.foldingathome.org/donor/id/8514

In firedfly 

The Corsair memory has arrived.  Thanks!

3 hours ago, neurotix said:

I'm afraid folding has killed my beloved 4090 Gaming X Trio, but it's done 2.3 billion points for the team.

 

Anyway TL;DR: folding has appeared to kill my 4090, will get a new temporary card and starting the RMA process with MSI.

Keep me updated on how the RMA goes.  I have the same card, folding 24/7.  No issues after 3 months folding at stock speeds.

in firedfly

in firedfly

37 minutes ago, Alex said:

Nice! Didn't realise how much more powerful hardware has gotten for folding, was surprised how my 'ancient' 1080 TI felt buried haha.

Just the difference between the 3000 series nvidia and 4000 series is significant.  The 3090 will get 8-9mil ppd, I believe.  My 4090 averages around 27mil ppd!  That is a HUGE jump in a single generation.

3 hours ago, ConsiderationOn said:

I noticed you're using vast.ai on your HFM Stats. Could you tell us more about it? Which docker image did you use? And how much is your budget for those 4090? 

I have a custom docker image that I created and a "startup script" to configure and start folding.  However, there is more configuration that needs to be done so you can manually stop folding after completing your last WU before you terminate the instance.  I'm hesitant to provided all of the information publicly for fear of people trying it out and having to terminate instances that have been assigned a WU.  That just slows down the science waiting for those WUs to timeout and be sent to others for completion.

If you (or anyone else) are interested in setting this up, let me know and I can help you get things configured.  Hitting me up on the EHW discord is likely the easiest way to get you up and running.

As for budget, I initially planned on $200 as an extra push for the competition.  Then @CoolGTX showed up in force .  I added another $200 to rent more 4090s to be able to compete with the massive amount of gear that he owns and can run!!!  In hindsight, it might have been better to use that money towards another 4090 to run full time  

2 hours ago, CoolGTX said:

@ firedfly

 

Your turn, any photos or details to share, I see what you did there.  Nice PPD jump

I guess a bit of background first.  For the last 1.5 years or so, I was running 3x3080TI and several lesser 3000 series GPUs.  Due to a recent move, I wasn't able to keep running all of that myself.  Two of the 3080TI's were shipped to @Cerberus to add to his folding farm.  He hosts cards for several EHW members.  Meanwhile, I picked up a 4090 that I fold on 24/7.  

For the competition, I had the following personally owned hardware in play:
- 1x4090

- 2x3080TI  (hosted by cerberus)

- 1x3080 12gb

- 1xQuadro P620  (This thing makes a crazy amount of PPD!!!)

I'm guessing I would have been fighting for third with my hardware alone.  As mentioned above, I also rented 4090s on vast.ai for the competition.  My initial ppd was with 4x4090s being rented (about 170mil ppd in total with my hardware).  After you ramped up, I added another 8x4090s.  I'm currently renting 10x4090s total and my combined ppd with my personal hardware is about 325mil ppd.

I'll try to add a few photos later tonight.

@CoolGTX, from my understanding, you have only used hardware that you own and run.  That is an amazing amount of hardware to be able to devote to folding.  You have put up MASSIVE points during the competition.  Thanks for contributing and I'm glad we were all able to push each other to contribute more than normal during this competition!  A lot of science has been done.

@CoolGTXstepped it up another notch and is closing the gap to first place!  What hardware are you running and what did you change for the extra push?

46 minutes ago, The Pook said:

 

haven't really been keeping that close of an eye on it but ~5-5.5M for the most part?

 

but half the time I'm limiting it to ~750mv and when I'm not in the room (aka getting cooked) I'm letting it run at 1.0v. it's lower than lar.systems says but I assumed that was from limiting the voltage.

If you want to spend any time trying to optimize those 3080ti's for the best PPD, let me know.  I think we could bump those up a bit.  @Cerberusis currently running my 3080ti's and reported earlier today that they are between 6.5-8.5mil ppd depending on the WU.  I think he's running them in Linux.

By comparison, my 3080 12gb in windows is doing ~7.5mil ppd pretty consistently.

39 minutes ago, The Pook said:

 

ETA is now 21 hours 32 minutes, aka in ~19 hours the ETA dropped ~9 hours 

 

wish folding.lar.systems had more complete data, curious how much faster a 7900/7950X or 13900K would be on the same WU. 

 

What's the ppd on your 3080ti?  From my (limited) testing, the drop in ppd on a high end GPU was more than a CPU would make.  So, folding on GPU+CPU was less ppd than just GPU.

There is no performance improvement with v8 and it is still beta.  If you use hfm to monitor your folding systems, it will not work with v8.  I'm not sure if any paths change, but I think they probably stay the same.

I wouldn't recommend using the v8 client right now unless you have a specific reason.

On 26/02/2023 at 22:32, Cerberus said:

That 5 gig is pretty cheap.  That's ~$180/mo here.

1 minute ago, damric said:

I think Q-cruncher is @Supercrumpet

Q-cruncher is @tictoc.

@Supercrumpet is cptasian.

In firedfly

On 01/02/2023 at 10:34, Avacado said:

Contributed 10% of your goal! GL reaching PCIe lane bliss. TY for all you have done thus far my friend!

I matched @Avacado's contribution, so you have another 10% from me.  Thanks for hosting!

in firedfly

My p620 was down for a bit at the end of the month due to some network changes and not having a monitor easily accessible to hook up to that system.  This has been resolved and the p620 is folding again.

1 hour ago, keeph8n said:

Thanks for letting me play. Ready for the next one

Sounds like you'll be difficult to beat next time!

24 minutes ago, neurotix said:

"All "Folding Time" prizes will require your username/passkey to be given to the donator, this will not give them any of your personal information, it just let's them be able to contribute full points to your username.  Passkeys need at least 10 full Work Units to start getting the full Quick Return Bonus.  Alternatively, you could setup another passkey for your Folding@Home username with a different email address and fold your first 10 units on that.  This would still put the points towards your username, but they would also be tracked by a different passkey.

Please note all winners must claim within 30 days from announcement of the winners."

 

I'm confused by this. Is there anything I need to do like set up a new passkey and change it and use it only for those 72 hours, or can I simply continue to fold as I have been and you'll track my points when I start and when I finish? I've never done a foldathon. Also, my current passkey and username were made on an email I cannot access anymore, thankfully my wife had an old email with a copy of my passkey I've used since I started folding in 2012.

 

Thanks.

You do not need to do anything different.  Keep folding with your current username and passkey.  The section you quoted is for when there are prizes awarded where someone will fold under your name for a period of time.  For example, last year, I offered up a folding time prize so someone could receive all points from a 3080ti for 24 hours.  The winner was randomly chosen from all contributors in the competition.

In firedfly

18 minutes ago, neurotix said:

Earned 22m points for the team so far and it hasn't even been 24 hours. lol

 

What ppd estimate are you seeing on the 4090?

Bowling

@Cerberus I finally got my Quadro P620 running again.  You shouldn't be the only one on the team putting up points now!

16 minutes ago, Avacado said:

360mm top/bottom is easily attainable in the O11XL. I am not sure on the mini. 

I was looking at the O11 Dynamic Evo.  It looks like its a few inches shorter than the XL, but similar length and width.  I wasn't looking at the mini.  The O11 Dynamic Evo should fit well in the space I have.  I could probably move to the XL if I needed to (I have more space for a computer than I expected I would when I started this thread).