Posts by Mr. Kevvy


log in
41) Message boards : Number crunching : Average Credit Decreasing? (Message 1792943)
Posted 22 days ago by Profile Mr. KevvyCrowdfunding Project Donor
1 Credit or 1,000,000 RAC. Wat's Da Bigggeee?


I can live with it (though it's kind of a pain) but there are volunteers either leaving the project over it or aborting GUPPI work units by the thousands, neither being good for the long-term. I can empathize as they are/were spending not only thousands on hardware, but for many similar amounts on power to do a certain amount of work. And now suddenly that is cut in half...

As well, as SETI@Home has been tapped to do the processing for BreakThrough Listen, there is now too much work available, but this work is being done at roughly a third the speed it should/could by volunteers' GPUs. If this bottleneck could be reduced or eliminated, we may be able to keep up.

OK, got the obvious out of the way. :^)


I don't care if it's me, Petri, Raistmer, or someone else entirely. Everyone involved in some small way will have to do the 'jiggly worm EUREKA dash'


So it turns out to be a software issue after all. :^D
42) Message boards : Number crunching : Average Credit Decreasing? (Message 1792891)
Posted 22 days ago by Profile Mr. KevvyCrowdfunding Project Donor
I had a look at 30-40 AMD GPU machines' tasks, and as far as I can tell, it looks as though new AMDs are similarly affected. I was unable to find any that didn't take 2-4x as long to complete a GUPPI VLAR work unit. This includes with and without SoG. Maybe I just didn't look in the right place, so if anyone has any URLs otherwise they would be welcome.

Old AMDs I saw were very slow but they were consistently very slow with the VLARs, and this would reinforce that it's an issue with the VLAR results exceeding cache size, as the old cards would have hardly any cache so both VLAR and non-VLAR work would exceed it so have about the same runtimes.

If correct, this would seem to imply that the only resolution for this is going to be a workaround redesign of the pulsefinding algorithm, if this is even possible.
43) Message boards : Number crunching : nvidia_opencl taking 100% of a core? (Message 1792741)
Posted 23 days ago by Profile Mr. KevvyCrowdfunding Project Donor
A few threads down: GPU tasks taking 100% cpu.
44) Message boards : Number crunching : Windows 10 - Yea or Nay? (Message 1792583)
Posted 23 days ago by Profile Mr. KevvyCrowdfunding Project Donor
There is nothing wrong with old technology if it works as designed and does its job. If you over-complicate with the "newest and latest and greatest" you introduce the potential for bugs and errors and security holes and malware, and the more components a thing has, the more there is to fail.

Example: homeowners who weren't happy with a simple mechanical thermostat that works for decades, and needed a Nest "smart" thermostat connected to the internet. Except now it goes and updates its firmware without asking anyone, and the update fails, and it doesn't work anymore which means the furnace is off and perhaps the water pipes will freeze and burst when the owner is away... now imagine this is the cooling system of a nuclear power plant instead.

Some of these "old tech" systems were engineered to a much higher standard of uptime reliability and real-time operation than any consumer-grade OS. Replacing them with a general-purpose machine could compromise this, ie "Sorry... can we put off the imminent attack for an hour? Windows is upgrading again."
45) Message boards : Number crunching : Average Credit Decreasing? (Message 1792091)
Posted 25 days ago by Profile Mr. KevvyCrowdfunding Project Donor
. . But is that option to refuse all GPU based Guppi WU's? It would be effective but a little drastic.


That may be, but it would offer an alternative to those who are aborting GUPPI work units by the thousands, which is far more drastic. And since the SAH team is apparently working on it, it's "approved".
46) Message boards : Number crunching : Average Credit Decreasing? (Message 1792074)
Posted 26 days ago by Profile Mr. KevvyCrowdfunding Project Donor
re:
a solution to allow GUPPI VLAR GPU work units to be refused in the Project Prefs. is being worked on...


Since I posted this, I've been more and more mystified by one facet of it:

It was indicated that this was tried on Beta and the result was no GPU work at all, which is why it isn't already implemented here. However, the code to do this here was already in place and was working properly before (as we had no GUPPI work on NVidia for weeks)... it just needs to be put back to the way it was, with one added check for the participant's profile setting added.
47) Message boards : SETI@home Science : Kepler Mission Discovers 1,284 New Planets! (Message 1791194)
Posted 28 days ago by Profile Mr. KevvyCrowdfunding Project Donor
I hope that this one (Kepler-62f) is being looked at...
48) Message boards : Number crunching : Average Credit Decreasing? (Message 1791142)
Posted 28 days ago by Profile Mr. KevvyCrowdfunding Project Donor
Perhaps it's just my ignorance showing again. but to somewhat improve the situation, would it not be possible to undo the misassignment of GUPPI VLARs to GPUs while CPUs are getting Arecibo MB tasks?

It appears that the file that controls this is sched_request_setiathome.berkeley.edu.xml in the project folder. As well, there are boinc_task_state.xml state files while they are processing.

GPU-assigned work units show:
<app_version>1</app_version> <plan_class>opencl_nvidia_sah</plan_class>


whereas CPU-assigned show simply:

<app_version>0</app_version>


It would appear that if this file were loaded, instances of each counted and then switched if there were both GPU "guppi" units and non-guppi CPU units (as the count has to remain the same of each... 100 max. per CPU and GPU), ensuring that they weren't present in any of the boinc_task_state.xml.

Problem is if this was being done constantly we'd wind up in the same situation because eventually they would be all slow guppis, the faster non-guppis having been reassigned to GPUs and completed. So, maybe every few hours would be ideal.
49) Message boards : Number crunching : GPU Wars 2016: News & Rumors (Message 1791000)
Posted 28 days ago by Profile Mr. KevvyCrowdfunding Project Donor
That would be excellent and thank you. Stock is fine so whatever the project installs automatically, as there's no known optimization specifically for the VLARs currently anyways.
50) Message boards : Number crunching : GPU Wars 2016: News & Rumors (Message 1790987)
Posted 28 days ago by Profile Mr. KevvyCrowdfunding Project Donor
If anyone obtains a 1080 (or even a 1070) and starts using it on this project, please let us know here. I'm interested (and probably many others!) to know whether the changes in the Pascal architecture mitigate the VLAR issue at all. If your computer(s) aren't hidden, we can have a look.
51) Message boards : Number crunching : Windows 10 - Yea or Nay? (Message 1790326)
Posted 25 May 2016 by Profile Mr. KevvyCrowdfunding Project Donor
Microsoft has backtracked on the "clicking the 'X' indicates acceptance" upgrade notification

Microsoft has u-turned over changes it made to a pop-up encouraging users to upgrade to Windows 10.

Users were angry that clicking the cross to dismiss the box meant that they had agreed to the upgrade.

Based on "customer feedback", Microsoft said it would add another notification that provided customers with "an additional opportunity for cancelling the upgrade".

The pop-up design had been described as a "nasty trick".

Microsoft told the BBC it had modified the pop-up as a result of criticism: "We've added another notification that confirms the time of the scheduled upgrade and provides the customer an additional opportunity for cancelling or rescheduling the upgrade.
52) Message boards : Number crunching : Decided to Bail Out on SETI for a Short While (Message 1790131)
Posted 25 May 2016 by Profile Mr. KevvyCrowdfunding Project Donor
The symptom is abandoning work because of the of the flawed credit system.


As noted too often (including in point #5 below) the people who are doing this inevitably have NVidia cards, which are processing the GUPPI VLAR work slowly due to a design flaw in the memory architecture. Unfortunately, the credit system won't fix that. The only way of doing that would be to have a special multiplier to arbitrarily grant 2.5x the credit for a GUPPI VLAR work unit if it's done on an NVidia card. If that is done, then why shouldn't every other sub-optimally slow platform get its multiplier?
53) Message boards : Number crunching : Decided to Bail Out on SETI for a Short While (Message 1790125)
Posted 25 May 2016 by Profile Mr. KevvyCrowdfunding Project Donor
In case it was missed in the decreasing credit thread:

I was able to contact Dr. Eric Korpela. He indicated that a solution to allow GUPPI VLAR GPU work units to be refused in the Project Prefs. is being worked on. Please pass it along...

Hope this alleviates any of the misgivings that the project admins. aren't aware of this issue or don't care about it.


They know. I'm in the same boat... I just spent thousands to upgrade the farm to all Maxwells plus a new four-card cruncher and this happens. I try to look at the bright side. Here's the bright side:

1) The work we're doing is far more likely to find something, being targeted at known nearby planetary systems with planets as similar to ours as possible.

2) We've gone from a work shortage to too much work; there's always enough now.

3) Because this project is being scrutinized as part of Breakthrough Listen, this work is going to be used, analyzed and searched. It is not going to sit in a database collecting digital dust. I can't tell you how I know this, but I do. This work is worthwhile.

4) My GPUs don't run dry of work on short outages (ie Tuesdays) because the VLAR work units take longer.

5) This problem is due to the memory architecture of NVidia cards. It isn't the SETI@Home team's fault, so at least there's that. There's no malice or sloppy coding etc. behind it. If anything we should complain to NVidia as they are pushing CUDA as a scientific computing platform.

6) The GPUs run cooler and so will probably last longer.

7) Half the world can't even get enough food to eat, so we're pretty well off if this is all we have to worry about. Gotta mention this. :^)
54) Message boards : Number crunching : Average Credit Decreasing? (Message 1789620)
Posted 22 May 2016 by Profile Mr. KevvyCrowdfunding Project Donor
I was able to contact Dr. Eric Korpela. He indicated that a solution to allow GUPPI VLAR GPU work units to be refused in the Project Prefs. is being worked on. Please pass it along...

Hope this alleviates any of the misgivings that the project admins. aren't aware of this issue or don't care about it.
55) Message boards : SETI@home Science : SETI E.T. Search Radio Interference ?? (Message 1789591)
Posted 22 May 2016 by Profile Mr. KevvyCrowdfunding Project Donor
Radiotelescopes are usually located in radio-quiet zones to prevent interference, ie Green Bank's location was chosen as it is surrounded by hills. There are special restrictions in place ie no cell phones, diesel vehicles only (no spark plugs). Any machine noise in the area would be either below the horizon of the telescope or not be concentrated by the dish; it would have to strike the receiver directly and thus, at such a distance with inverse square propagation would be very weak perhaps even below the noise threshold. That frequency range doesn't bounce off the ionosphere either so wouldn't hit the dish by reflection.

So, I don't think it's much of an issue. I'm sure it would have been taken into account! :^)
56) Message boards : Number crunching : Average Credit Decreasing? (Message 1789560)
Posted 22 May 2016 by Profile Mr. KevvyCrowdfunding Project Donor
Since he is a member of the GPU Users Group perhaps some one with diplomatic skills could PM him and explain to him how is is screwing up SETI.


I'll pass the word. I thought it was this machine's owner but it turns out this is someone else doing the same thing. Shows 824 aborted GUPPIs.
57) Message boards : Number crunching : Average Credit Decreasing? (Message 1789284)
Posted 21 May 2016 by Profile Mr. KevvyCrowdfunding Project Donor
To add to what Chris wrote: at first there was much complaining that there was never enough work.... we were running out every Tuesday and not catching up until the next day, and plenty of other times as well.
Solved: the SETI@Home scientists built the GBT receiver and a new splitter and got on the Breakthrough Listen initiative, and now we have enough work (possibly even too much!)

Next, there was much complaining (including from myself) that CUDA GPUs weren't being assigned the new GBT work, though it's possibly far more likely to contain a candidate signal, and thus find something.
Solved: the same group beta-tested the impact to system stability/usability of these work units being run on CUDA GPUs and then released them (this was also probably necessary, because there is so much work and CUDA GPUs make up such a significant fraction of machines available to do it.)

Now, there is much complaining that this work is causing a reduction of credit. Sure, it's sad that our RAC is dropping, but it isn't anyone's "fault" except the NVidia engineers who designed the memory architecture of the CUDA-enabled cards, and it's the unintended side effect of something that was asked for. It isn't CreditNew or Dr. David Anderson's doing. Fixed credit won't solve it as 3x the complete speed still equals 1/3 the fixed credit over time. It needs a possibly complex, and possibly impossible, workaround in the SETI@Home client itself.

So, in the interim, be patient, enjoy the new science, and be happy that those GUPPI work units trade off three times the compute time for... I dunno... thousands of times the chance of actually finding something, and I'll take that tradeoff any day!
58) Message boards : Number crunching : Welcome to the 17 Year Club! (Message 1788987)
Posted 20 May 2016 by Profile Mr. KevvyCrowdfunding Project Donor
Missed mine five days ago... better late than never. Story of my life. :^)
59) Message boards : Number crunching : Average Credit Decreasing? (Message 1788817)
Posted 19 May 2016 by Profile Mr. KevvyCrowdfunding Project Donor
I'm just crunching for Milkyway for a week and already have a higher RAC, than i had here before...


And if I move to Mexico and get a job as a janitor my paycheque in pesos will be larger than it is currently in dollars.

There won't ever be a BOINC consistent credit payment across projects, because smaller or startup projects or those with not-so-interesting (or nigh-incomprehensible) goals overpay to attract volunteers, and will continue to do so.
60) Message boards : Cafe SETI : Thor is the Designer Cafe! (Message 1788769)
Posted 19 May 2016 by Profile Mr. KevvyCrowdfunding Project Donor
Yes Thor wants enthusiastic support.


Enthorsiastic!


Previous 20 · Next 20

Copyright © 2016 University of California