Hey

[Nuadormrac]

They've passed AMDusers on POEM according to BOINCstats, so I've moved POEM over here now... As of yesterday, for whatever reason (not sure if it was a change at POEM in the WUs, ramping up on SETI for the wow signal thing, or whatever, but the comp is running a lil hotter now, so the GPU has lowered itself to the base clock of 745 MHz core clock on the GT 650m, 1 GB gddr5, vs the boosted clock of 835 MHz it was running at before. It has drawn out crunching time from 28 min to about 33 min on most POEM tasks. Other GPU projects, after things have had some time to cool off from not crunching POEM in awhile go back to 835 MHz crunching... Was something yesterday, just hoping it wasn't having turned on the GPU at my mother's apartment, where the a/c in her apartment is a lil flakey...

The stat sites should catch up wrt where the credits are going within the next day or 2....

[Crazybob]

Welcome to the team. If I remember right, you need to change your preferences at FreeHal to only accept the low CPU usage WU's. That way it shouldn't eat up your memory. I could be wrong though.

[Nuadormrac]

Yeah, thx, not sure what was going on with that, as I was already set to only use non-CPU intensive tasks... I think 8 FreeHAL tasks, + some other stuff, really pushed it, when projects were sending, considering my old laptop only had 2 GB RAM (though a core 2 duo), and this has 4 hyper-threaded cores (or 8 as seen by Windows/BOINC), I didn't exactly expect that one, lol

[Mad Matt]

Welcome back Nuadormrac, and thank you for the support. If you still are having problems with tuning the GPUs in, just let us know.

[Nuadormrac]

Well yeah, there's a dual side to this, though the past week I was pushing more onto SETI (normally don't run it much anymore though) due to the current anneversery event. Though an odd one, I might wait to try getting a second task on. OK, a little background here...

I recently got a new laptop, and as I need a computer loan at the present time (stinkey economy, I could afford to make payments, but not a whole new system outright, also one of my jobs is seasonal, return to work in a week, and then 2 for a paycheck) I got an Alienware x14 r2. For some things I do need a portable computer, and I only build desktops, not laptops. I'm not even sure one could get a supply of standardized components to custom make a laptop, it's not like it's in a standardized form factor such as ATX. Anyhow, Dell has financing available, so...

Only thing, the cooling on the thing is such that the CPU and GPU share cooling pipes, so when something drives the temps up, some of it translates to the other, aka the CPU runs hotter then the GPU in this. Not too bad, except last Sat, well hmm... Needless to say, I had to print something, the a/c in my mother's apt isn't what it should be, and yet without starting the GPU the CPU temps were below what it would normally run in a room more around the temps I like. I loaded up SETI, and then POEM was having probs where they pulled a few tasks, and the GPU throttled it some. All seems fine, but I'm not sure if for now I want to push a second POEM task on there (oddly enough SETI CUDA runs hotter, but it has no such down clocking of the tasks) until I get a cooling mat for this thing. Though all seems OK, I was scrambling to figure out wtf was going on, as POEM tasks were taking a bit longer, and some I guess weren't GPU crunching at all. I think some of their tasks had issues; but in either case, I'd rather be safe in that regard. After I get a cooling mat, so I don't have to worry about the thing overheating (the CPU had hit 85, 86 C, which is getting close), I'll look at throwing a second on there....

Anyhow, all the configinfo.xml files I saw for POEM are ATI specific, but the card I have is nVidia, so when I'm ready to go ahead, it's a matter of something that works with nVidia cards. I've got a Keppler based GT 650m, 1 GB gddr5 here, coupled to an i7 quad, using the Ivy Bridge....

Well after being left to wonder if something happened starting up the GPU in my mother's rather warmish appt, to then finding some oddities on the project (though they've released new WUs since, after having gone dry, which do yes, take a touch longer to complete), I'll probably hold off till next month and then make the changes, after I can purchase a lil more cooling for this thing... It's brand new, and all of that...

[Nuadormrac]

Ya know? I don't think SETI was ready for a team to host an event, getting the other teams to crunch SETI in a bit of a competition. They had the one planned maitenance, OK, it happens. And prevenative maitenance can be a good thing. But their project is back down, after large scale difficulties downloading WUs from people around....

Obviously, the wow signal event is related to SETI, for quite obvious reasons; but I think some other projects can handle a ramp up in production a little bit better at present. Might be that they're not as big to begin with, but if someone were to host say a competition and get a fair few of the teams to participate in say Climate, would they have quite the same number and frequency of server shutdowns and BW/server issues people are seeing getting SETI tasks now? Not sure I've ever really seen Climate go down, except the forums once or twice...

[BlkJack-21]

Welcome to the forums Nuadormrac