Thanks for your reply. I don't think they answered the questions I just posited. Maybe somebody in the establishment would kindly take a shot at just that. If no one answers, then I assume the answer must be that the establishment has decreed that they want to artificially restrain the amount of work put into astropulse.

I'm not positive that the motivation is just because there are more points involved. Maybe some people have decided based on the scientific merit of pursuing astropulse.

How much data could be contained in one of these pulses? How far out in light years is astropulse 'good for'?

Thanks for your reply. I don't think they answered the questions I just posited. Maybe somebody in the establishment would kindly take a shot at just that. If no one answers, then I assume the answer must be that the establishment has decreed that they want to artificially restrain the amount of work put into astropulse.

I'm not positive that the motivation is just because there are more points involved. Maybe some people have decided based on the scientific merit of pursuing astropulse.

How much data could be contained in one of these pulses? How far out in light years is astropulse 'good for'?

What an overly cynical conclusion based upon no information whatsoever.

The reason why there's always a shortage of AP workunits is because they're cut from the same "tapes" and they're much larger, so there's less per tape than MB workunits. MB is approx. 365KB in size. AP is approx. 8MB in size. Each "tape" is about 50.20GB in size. Ergo, less AP is created than MB from the same tape. There is no artificial restraining of workunits for any other reason.

BTW - I keep putting tapes in quotes because they don't use tapes anymore, so it has become sort of a colloquialism. The tapes are in fact hard drives now.

Thanks OzzFan. So now I know and don't have to be so overly cynical anymore
and can base my views on actual information.

How much data could be contained in one of these pulses? How far out in light years is astropulse 'good for'? [This forum is so good, that I wanted to ask these questions here rather than on the science message board. If that is OK.]

How much data could be contained in one of these pulses? How far out in light years is astropulse 'good for'?

I honestly don't recall such information. I would strongly suggest asking in the Number Crunching forum where Josef Segur posts frequently. I'm certain he would know the answers to your questions.

Zalster:
My 3 gpu's are now running fine and I have no trouble getting astropulses for them. I would now like to add a fourth. I have a PCIE x1 slot available. Can I somehow run a PCIE x16 card on it? I guess I need some kind of riser and then have to buy a half height card and cut it down to fit?
My power supply is fine. The loaded system I have now only draws 320 watts and I have 600 useable watts available.

If that other card can sit on your motherboard in that PCIe slot and you have the free pin connectors for it, then yes you can use it. The only time I used risers were was when I didn't have an available PCIe slot. But you have to be careful if you ever decide on one of those. I'm not going to go into detail unless you decide to go that route. For now, if you can sit another GPU in a slot and power it, then you can use it. The throughput for seti is very low so a x4 is more than adequate.

Your thinking is correct. Normally you have 1 or 2 slots at x16, but there are other slots sometimes as well, when you connect to those other slots, it lowers the throughput to the original so then they become x8 or even x4 if you are connecting multiple slots. It gets pretty technical to explain. Let me find a link for you


http://www.enthusiastpc.net/articles/00003/

Not the original article I wanted but close enough.

Happy Crunching..

Zalster

The issue of using a riser is 2 fold. Do you have the power to supply that extra gpu and do you have the space. When I was using my risers, I had to figure out where the GPU was going to sit or hang from inside my case. I was fortunate that my PCIe x1 as low enough on my MB that I could sit the GPU on the bottom of my case. In another case and motherboard it was up high near the CPU such that I had to "hang it" from the case with string or wire so that it wasn't just lying on another GPU.




To the power issue, you are going to want to get a pcie1 to 16 riser. Make sure that there is a molex connector with at least 2 wires from the molex going to a capacitor on the riser. The reason for this, is you need to supply at least the 75 W that would normally flow from the MB to that riser. Also, you want to protect the GPU from surges. This does not include another other connection that you may need for the GPU. Some GPUs require a 6 or 8 pin connector directly on them. The top of the line GPU require both a 6 and 8 pin connector. So depending on the GPU you will have 1,2 or even 3 power lines from your PSU to that Card. There are a couple of places to find Risers, I used both EBay and Moddiy.com I prefer the Ebay. The other thing with risers, make sure it has that little sliding plastic part on it that holds the GPUs in place. If you look at the pic above, you will see it at the end of the x16 end. It's almost center top of pic at the end of the black riser, it's tan color. Some makers don't put that on, so the riser can just drop off the GPU if you have it hanging somewhere.

I never recommend cutting down a GPU. There's no rule saying that the GPU can't be hanging outside of the case. I know many people that don't have theirs physically in the case, main thing is getting it connected and powered. You might have to reinstall the drivers, don't let windows install them.

Lastly, restart. Depending on your Motherboard and PSU, it will either restart and will see all the graphic cards....Or it will fail in which case, you might get a blue screen or an error code. I had 1 machine where I tried to add a second GPU and it immediately only beeped after restart. Nothing I did made it load windows. I removed the riser and it restarted and it loaded. I added riser and it beeped. Short story, too much for that motherboard.

Happy Crunching...

Zalster

Edit..

Check the length of the ribbon of the riser... Not too short

Zalster,

Thanks much for your info and time. I will be working on this project. It will take me awhile to figure it all out for my case and working with ebay. It looks like your posting was very complete and clear. Thanks again.

Further report:
As to whether I use 1,2, or 3 wu's per gpu, I have found for my machine it is best to just use 1 wu per gpu because of overheating. My gpu's just get too hot trying to put too much of a workload on them.

After further experimentation, I am using .33 in app_config for my gpu's instead of 1 (# wu's per gpu). BillBg I am learning.:-)

Zalster:
I bought my riser but I can't reach it to the floor of the box. I can put it up higher over the area of the hard drives. I have a bunch of unused cables from the power supply that overlay the hard drives. It would make a nice 'bed' for one of my old 6570's. Do you see any problems with this?

The issue of using a riser is 2 fold. Do you have the power to supply that extra gpu and do you have the space. When I was using my risers, I had to figure out where the GPU was going to sit or hang from inside my case. I was fortunate that my PCIe x1 as low enough on my MB that I could sit the GPU on the bottom of my case. In another case and motherboard it was up high near the CPU such that I had to "hang it" from the case with string or wire so that it wasn't just lying on another GPU.

Zalster

Edit..

Check the length of the ribbon of the riser... Not too short

I probably wouldn't lay it on top of the cables. Maybe you could find something that your could put over the cables? I had a piece of hard clear plastic that used to use to insulate the GPU from other parts. What about one of the hard drive harness to keep it from touching the cables? or can you lay it in an empty harness? just thinking of possible options.


Zalster

How about if I used the anti-magnetic envelope that the riser came in to shield it from touching the cables? Thanks.

No, I wouldn't use that. I'm more concern about heat. Some GPUs can get really hot, so you don't want them just lying on exposed wires (even insulated ones). At one point I "hung" a gpu from one side of the case using thread thru the hole in the face place (where you would tighten the screw to secure it to the case) up to the top of the case. Not pretty but it worked. The hot air was pulled up and out of the my case so no build up. If you can find something that is heat resistant to put the GPU on then you can use that just as well. My 2 cents.

Zalster

OK, thanks. I'll work and see what I can come up with.

Zalster:
I have another PC where it would be easy for me to place the riser on, however it only has PCIE 1.0. Would that work? I have enough watts available on that one too. My graphics card I currently use on that one is in the 4000 series and only does integer math. If not, I'll work something else out.

No, I wouldn't use that. I'm more concern about heat. Some GPUs can get really hot, so you don't want them just lying on exposed wires (even insulated ones). At one point I "hung" a gpu from one side of the case using thread thru the hole in the face place (where you would tighten the screw to secure it to the case) up to the top of the case. Not pretty but it worked. The hot air was pulled up and out of the my case so no build up. If you can find something that is heat resistant to put the GPU on then you can use that just as well. My 2 cents.

Zalster

Hey Merle,

It should work, that is what I had my other risers connected to in those machines.

Zalster

Thanks again - will do!

One 6570 gpu has been throwing out errors for past hour or so. I have two 6570's and one 7770. The temp on the one throwing out errors has been showing a low temp reading lately. First off, I don't know which of the 6570's is the faulty one. How do I go about diagnosing the problem? Do I have to catch the std err file after uploading and before reporting? I have to stop my operations now because of so many errors.

Can I shut off using gpu's at my seti preferences without losing all of my gpu wu's that I have already downloaded while I straighten out the problem? I want to do this in order to continue using my cpu's.

Merle,

All the reported errors so far on your computers are coming from "aborted by user". Let them upload and we can look at the stderr after they upload. Right now there is no way for us to tell what the problem is until then

Zalster

Zalster,
They are being shown now as inconclusive results. Thanks

Inconclusives are not Errors. That just means that they don't match up close enough to your wingman. They will be sent out again to a 3rd person. It will keep doing that till either there is a match or they determine that there is an error with your calculation. Either way, you have to wait and see what happens. Having the occasional error is ok, it's when you start to have like 20-30 or more that you need to look at your system and see where the problem is. So until then, just keep an eye on it.

Zalster


Edit.. Looks like they all came from the Device 0 Look at your advance log from start up and see which device it is. I'm guessing it's 1 of the 2 identical one's you are using. The 3rd has a different name to it. Are you using the riser in this machine? If so, what GPU is it connected to?