Posts by Richard Hagen

1) Message boards : SETI@home Science : Arecibo Observatory (Message 2062683)
Posted 1 Dec 2020 by Profile Richard Hagen
Very sad news. I would like to see it enlarged, modernized and rebuilt. Arecibo still maintained some unique capabilities, even with the larger single dish antenna now online. We have lost the most powerful radar telescope and planetary radar we had, and we have lost some of those abilities for the time being at least. I believe that the facility was good for Puerto Rico and the U.S., and in fact the whole World. Look at all the data that we have crunched that was captured by that very telescope. A sad day. It was inspirational, and we need to rebuild it, but bigger and better. Rich
2) Message boards : Number crunching : User achievements thread......... (Message 1924396)
Posted 14 Mar 2018 by Profile Richard Hagen
Just made 10 Million. Working on a multi GPU rig for the fun of it, although I am not sure how to put all of that waste heat to use now that it is starting to get warmer out. Rich
3) Message boards : News : No work. Database issues. (Message 1904471)
Posted 2 Dec 2017 by Profile Richard Hagen
I am glad it is back up and running. I would echo Dahl's comment as I have not turned on any supplemental heating yet. Fortunately it was warm the last couple of days in Chicago. I worked on a few other projects, but the lack of the heat generated crunching SETI data was actually quite noticeable. Rich
4) Message boards : News : No work. Database issues. (Message 1904034)
Posted 1 Dec 2017 by Profile Richard Hagen
It looks like I have started receiving new jobs now. It is good as my faster rigs and even some of my not so fast had run dry. Rich
5) Message boards : News : No work. Database issues. (Message 1903690)
Posted 29 Nov 2017 by Profile Richard Hagen
Thanks for the information, Some of my machines had gone idle and I could not get them to load new jobs and I had been thinking that the problem might be on my end. Appreciate the work you guys are putting in to keep the servers functioning. Rich
6) Message boards : Number crunching : SETI/BOINC Milestones [ v2.0 ] - XXIX (Message 1894834)
Posted 12 Oct 2017 by Profile Richard Hagen
Congrats to all on their accomplishments, I have just passed a million total in Seti, it has only been 18 and a half years or so since I joined. Rich
7) Message boards : Number crunching : Anyone ever play around with adding external GPU's to your laptop via the mini-PCIE? (Message 1889892)
Posted 15 Sep 2017 by Profile Richard Hagen
I have a some similar PCIe connectors for an external graphics card, but for a regular PC as opposed to a laptop. I have some concerns about the power supplied to the board for the video card.

A couple of things to check would be that the SATA specification calls for three +12V pins on one end of the connector. The pins have to be rated for a 1.5 amp minimum, but that is only 18 Watts per pin at 12V, if not all pins are connected it could be less on some inexpensive aftermarket parts as it seems a few do not connect the +12V precharge pin. Even with three +12V pins with the amperage split evenly, that is only 54 Watts. I suspect that you can probably get away with a bit more current before it melts down in real world conditions, but if the voltage to the card drops across the supply lines I would expect erratic results.
SATA specs are outlined here:

Now the 6 pin PCIe connector on the other end is supposed to be rated at 25 Watts or just over 2 amps per pin.
A source for the pinout and power specs is here:
The specification calls for there to be pins, connected to the two yellow wires, of +12V, although I have seen some with a connected third +12V wire in the not connected pin of the specification. If only two of the yellow +12V connectors are connected that would make the max about 50 Watts. If three of the pins are connected to +12V on BOTH the board connection and the cable then that would raise to 75 Watts. The kit your referenced appears to only have two +12V yellow wires. For 8 pin PCIe power connectors, the same amps per pin applies but you have three pins minimum, or 75 Watts.

So in short, it seems to me that there is a mis-match between the available power from the SATA end and the PCIE end. Depending upon the graphics card you employ, I am thinking you could easily exceed the power specification of the SATA end.

For example the GEForce GTX 980 Ti information linked through here indicates a power consumption of up to 250 Watts: The card I looked at has two 8 pin power plugs. If you were drawing 75 watts per PCIE connector you could pull in 150 Watts, through those connectors if properly wired. That would leave 75 watts to be pulled through that SATA power connector in your kit and/or the USB type cable.

There is that additional connection, the USB cable used between the two cards in your kit but I do not know the specifications for that in this application or if any of the lines are used to transfer power from the laptops PCIE rail to the rail on the board that the graphics card is plugged into it or if it is all data traffic.

For me, with a laptop, I would probably obtain an external +12V supply with the ability to put out the total wattage of the card and tie it into the board PCIE power and the cards PCIE connectors and placing the ground in common with the laptops for reference. I suspect that power problems may cause these kits as supplied to not work with many cards on the market.

Take all of this as just my two cents worth and double check my figures, as I have not rigged up mine yet either, but I would make sure that I have enough available for the card for it to operate, especially in such an intensive manner as this application.

8) Message boards : Number crunching : Good to be alive, here in Naples, FL (Message 1889887)
Posted 15 Sep 2017 by Profile Richard Hagen
I have friends on Marco Island, I was expecting the worst, watching from afar, but I did hear from them and they fared better than I had expected. Hoping the best for you down there. I guess you have your power back as well. Went to Naples a couple of times, I like the area. Last time I was there went to a place Shea's at Landown St., I hope it is still there. Rich
9) Message boards : Number crunching : Older 8Gb Galaxy Tab 2 units will not take new jobs (Message 1889506)
Posted 13 Sep 2017 by Profile Richard Hagen
I can't see how to get to a 'cc config' file to edit it. I ended up reinstalling BOINC on them all and tinkered with the space allocations in the preferences again. I do not expect it to work. One other change that could have effected this is that I did go into the preferences a while back and adjusted them to use 2 processors instead of the default 1 processor. When they stop taking new jobs again I think I will set a couple of them back to using one processor and see if that has any impact on the problem.
10) Message boards : Number crunching : Older 8Gb Galaxy Tab 2 units will not take new jobs (Message 1889479)
Posted 13 Sep 2017 by Profile Richard Hagen
When I get back home I will try and find the file to mod and fiddle with the computing preferences some more. They had been working right for a while and then stopped, so I am thinking something changed, but I don't really understand what is going on with it at this point so I have a bit of learning to do.
11) Message boards : Number crunching : Older 8Gb Galaxy Tab 2 units will not take new jobs (Message 1889441)
Posted 13 Sep 2017 by Profile Richard Hagen
Someone here may know the answer to my problem. I have a bunch of old tablets from when I used to replace broken touch screens on them that I had working on Seti@Home and they were churning along fine, finishing a loading new tasks. Now some of them are exhibiting an error and not loading new work units. This all started around the time of the doubling of the work unit size.

The Samsung Galaxy Tab 2 unitsunits still working are the "samsung SAMSUNG-SGH-I497 - SDK:16 ABI: armeabi-v7a" units and the "samsung GT-P5113 - SDK:17 ABI: armeabi-v7a" units which all have 10.1" screens. Again these are still functioning fine. These are 16Gb units.

In contrast the Galaxy Tab 2 units "samsung SCH-i705 - SDK:16 ABI armeabl-v7 a" as well as the "samsung GT-P3113 - SDK:17 ABI: armeabi-v7a" units all, to a unit, began reporting an error. These are all 7 inch tablets. I can see the error message under the projects option in BOINC for Seti, which states "SETI@home v8 needs 32.00MB more disk space, you currently have 0.00 MB available and it needs 32.00 MB." These are all 8Gb units.

My computers are set to be viewable.

I Checked and adjusted the disk space and ram settings to no avail, and the units themselves, most of which were factory reset and then had BOINC installed, still show 3 or more gigabytes of disk space free and a multiple of 32Mb in ram memory free as well. All have older versions of Android, but it does not seem to matter exactly which 4.x.x is on them and similar versions run on the 10.1 inch variants.

Adjusting the computing preferences does not seem to have any impact on this problem, although, uninstalling BOINC and reinstalling it will cause it to download and execute one or two work units before the error recurs.

I don't know if I am missing something simple or if there was a fundamental change that will prevent these units from participating in the project. I don't know where to go from here. These units are surplus and also not super productive, maybe 60RAC apiece when left crunching continuous, but they do not consume much power and I have a bunch of them. I had planned to keep these and another box of them I have churning out data as a better source of electric heat over the winter than just a space heater.

Another unrelated thing I am curious about is what type system gives the best ratio of calculations per Joule of energy consumed, but that's probably best left to a separate thread.


©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.