I recently noticed that Kevin https://setiathome.berkeley.edu/show_user.php?userid=35215 a long time top 20 member who had "dropped off" that range has changed from the 2990wx that he was running back to his 1950(w?)x.

Previously in private communication he had told me that he was no longer running Seti@Home exclusively. If you are not aware the Threadripper 2990wx cannot run all 64 threads and get any reasonable level of production due to the memory design problems of the 2970/2990 Threadrippers.
So he normally ran about 26 threads which allowed a robust level of production.

He had some kind of corrosion problem with his cpu cooler so has reverted to the 1950x, which he reports is running loaded on 90% of its 32 threads.

I am not sure if he is still splitting his projects or not. But he does say he has some bios tuning left to do to wring out the last bit of CPU performance. After all he is getting "only" 210,000+ RAC right now.

Tom

---

A proud member of the OFA (Old Farts Association).

Sounds like he was running one of the Enermax Liqtech AIO's that have well known corrosion failure problems.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Sounds like he was running one of the Enermax Liqtech AIO's that have well known corrosion failure problems.

No.

It is an XSPC Raystorm Neo, the base is pure copper, there was a build up of a black gunge mainly on one side of the block which I removed with a hot dilute vinegar solution and a fine bristled toothbrush.

The area that was under the gunge some of it looked as if some of the very fine fins were slightly worn or eroded.

I had a look as best as I could without stripping down the GPU's and could not see any other problems so just flushed the rest of the system a couple of times, swapped over the CPU and re-assembled the system.

GPU blocks are nickel plated and looked clean.

Due to work etc have not had a lot of free time recently:-(

---

Kevin

That is not good and caused by galvanic corrosion. A mix of nickel and copper in the loop probably the cause. Not normally a problem unless no corrosion inhibitors are used. Happens a lot in plain distilled water loops. See the issue at YT Bitwit's channel.
https://www.youtube.com/watch?v=isL20lm3i8c

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

That is not good and caused by galvanic corrosion. A mix of nickel and copper in the loop probably the cause. Not normally a problem unless no corrosion inhibitors are used. Happens a lot in plain distilled water loops. See the issue at YT Bitwit's channel.
https://www.youtube.com/watch?v=isL20lm3i8c

Nowhere near that bad, but I have now put a corrosion inhibitor on my shopping list.

System has been running on de-ionised water with just a silver coil.

---

Kevin

System has been running on de-ionised water with just a silver coil.

Now we have another reason that people suddenly stop contributing.....


They become "de-ionized"......








ROFLing.

Tom

---

A proud member of the OFA (Old Farts Association).

As a contingency plan if money gets tighter I will take this box https://setiathome.berkeley.edu/show_host_detail.php?hostid=8853145

Leave the 3 most efficient gpu's online and switch part of the cpu crunching to World Community Grid while taking ALL my other systems off line.

I was going to announce the possiblity of switching all my crunching to this system: https://setiathome.berkeley.edu/show_host_detail.php?hostid=8724359

But either the battery has croaked or something else. It has stopped booting :(

Tom

---

A proud member of the OFA (Old Farts Association).

I have switched the heavy gpu processing to https://setiathome.berkeley.edu/show_host_detail.php?hostid=8894290 (8 gpus) and am now running 1 gpu on https://setiathome.berkeley.edu/show_host_detail.php?hostid=8853145.

So the 2nd system will certainly drop out of the top 20. And could drop out of the top 100 even. The first system will take a while to get up into the stratosphere since it is starting from almost scratch.

Tom

---

A proud member of the OFA (Old Farts Association).

I have switched the heavy gpu processing to https://setiathome.berkeley.edu/show_host_detail.php?hostid=8894290 (8 gpus) and am now running 1 gpu on https://setiathome.berkeley.edu/show_host_detail.php?hostid=8853145.

So the 2nd system will certainly drop out of the top 20. And could drop out of the top 100 even. The first system will take a while to get up into the stratosphere since it is starting from almost scratch.

Tom

The high gpu box is up to 25th place.

---

A proud member of the OFA (Old Farts Association).

https://setiathome.berkeley.edu/top_hosts.php?sort_by=expavg_credit&offset=0
https://setiathome.berkeley.edu/show_host_detail.php?hostid=8837101

I want to welcome Kevin Olley back into the top 20. For "several" years his system was running a Threadripper2990wx 32c/64t)(@26 threads) with the 4 video cards he still has and was usually parked about 5th place.

Then he disappeared. PM conversation indicated he was crunching more non-Seti projects. Then a month or two ago I noticed he was now showing a Threadripper 1950x (16c/32t).
There was some kind of technical issue with the 2990wx so he reverted.
Considering he got maximum total production at about 26 threads on the 2990wx moving back to the 1950x Threadripper wasn't going to slow him down at all since it didn't have the butchered up memory model of the 2970/2990wx's.

I am a retired big truck driver. Kevin is still active in the UK as a line haul truck driver. I bought a TR 2990wx and learned what Kevin already knew. It peaks in production at about 26 threads on most computationally intense CPU tasks. I learned the hard way at 60+ threads more than a third to half of the CPU tasks were running slower than most of the CPU tasks I used to run on Netbooks (~6 hours)...

Tom

---

A proud member of the OFA (Old Farts Association).

Just MY off-the-top-of-my-head spin on "Why do people suddenly stop?"

Realize Seti@Home doesn't really "help" the world at all compared to other projects (it's just "credit" for credit's sake). i.e. other projects are more "worthy" (e.g. Rosetta?).

Carbon footprint.

Money.

Saw a therapist to help with their OCD behavior.

Lost access to work PC's they could "borg" for credit.

Many Seti work units are just a rehash of old units, what's the point?

Yeah, I'll probably get shunned/banned,

Yeah, I'll probably get shunned/banned,

Some may well shun you, but you are saying what you honestly believe, so it is their loss.
Banned - I don't see that your post breaks any forum rules so should you be banned.

Just keep on processing work as and when you want to, and can afford to; others should accept that we all have the right to do exactly that.

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

I have not lost faith.
I just cannot contribute at the level I did at one time.
Everybody knows that our Seti search is a 'needle in a haystack' proposition.
But if we do not search, our chances of finding anything are zero.

So I shall continue to contribute what I can to the project with the knowledge that at least we are trying.

Meow!

---

"Time is simply the mechanism that keeps everything from happening all at once."

it's also possible that one or many interesting signals might have already been discovered. they just haven't been looked at yet.

but the whole question this topic poses is strange to me. 17 out of the top 20 users are still participating... really the only anomaly in the list is Charles Long, who from what I can tell got his points load testing an entire datacenter for a couple years, then fell off.

being a 20 year old project, some people died, lost interest, had other priorities, etc.

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

I have not lost faith.
I just cannot contribute at the level I did at one time.
Everybody knows that our Seti search is a 'needle in a haystack' proposition.
But if we do not search, our chances of finding anything are zero.

So I shall continue to contribute what I can to the project with the knowledge that at least we are trying.

Meow!

+1

---

A proud member of the OFA (Old Farts Association).

I have not lost faith.
I just cannot contribute at the level I did at one time.
Everybody knows that our Seti search is a 'needle in a haystack' proposition.
But if we do not search, our chances of finding anything are zero.

So I shall continue to contribute what I can to the project with the knowledge that at least we are trying.

Meow!

+1

+2

Keep searchin'!
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

Many Seti work units are just a rehash of old units, what's the point?

1) An error in the Gaussian fit algorithm was corrected years ago.
2) The pulse-finding algorithm has been improved.
3) Autocorrelation detection has been added.

... and probably much more than this. So it is worthwhile to re-run old data through the system as there could be undiscovered signals hiding it in.

---

Many Seti work units are just a rehash of old units, what's the point?

1) An error in the Gaussian fit algorithm was corrected years ago.
2) The pulse-finding algorithm has been improved.
3) Autocorrelation detection has been added.

... and probably much more than this. So it is worthwhile to re-run old data through the system as there could be undiscovered signals hiding it in.

Also a lot of these old files were 1's that were pulled due to errors in them and last year they had an intern going through those files correcting them so that they could be split without the errors.

Cheers.

while I'm not at the top, and I never will be again, I did have (with permission) a bunch of work machines that I ran in the early days. I then dropped seti for a while as life changed (new home, bigger family) and didn't run seti for a while, when I came back I didn't merge my old account with this new account... so one of the old logins that doesn't produce anymore is my old one. I'm still doing seti, but I needed to take a break.

Thanks to Mr. Kevvy and Wiggo for the reasons running old data is useful.

... last year they had an intern going through those files ...

I wish that intern had left notes (or was still around to make notes) about what was on each tape. Having a flood of VHAR shorties cached to send out after a long outage this morning wasn't the cleverest move.