Personal computing discussed
Moderators: renee, farmpuma, just brew it!
Looking for Knowledge wrote:When drunk.....
I want to have sex, but find I am more likely to be shot down than when I am sober.
Looking for Knowledge wrote:When drunk.....
I want to have sex, but find I am more likely to be shot down than when I am sober.
farmpuma wrote:On my dual core cpus running the 30xx SMP WUs I noticed them missing all of the Preferred Deadlines. As Trax recently pointed out, when the WU exceeds the Preferred Deadline it is re-released for processing. I would be willing to wager that the majority of the cpus running the SMP WUs are dual cores. So, unless densign's objective is to cycle these WUs through as many systems as possible the excessively tight deadlines are highly counter productive.
On a semi related topic -
WU 2665 = WTF?!? .. They take nearly double the amount of time to process for 2/3 the amount of points of the other 26xx WUs. You want to do WHAT? with my hardware and electricity?
/end of rant
farmpuma wrote:Hewhoiswashin6x9s
Looking for Knowledge wrote:When drunk.....
I want to have sex, but find I am more likely to be shot down than when I am sober.
david00214 wrote:I'm getting a bit frustrated myself.
76 failed attempts! GAH!!!!
farmpuma wrote:Let's get this whine fest rolling ..
I apologize for my overly enthusiastic reaction to the points increase for the 2665 work units. At 1920 points they are still considerably less productive than any other SMP work unit, including the dreaded 3065s at 2144 points. Over the years I've done my fair share of crapola work units and would probably slog through the 2665s except for an additional major problem. The finished data file does not compress for upload like all the other SMP work units and for those of us still stuck on dial-up the resulting two to three hour upload is completely unacceptable.
jeffry55 wrote:
Double GAH!!
Do you get another WU while you wait to send?
david00214 wrote:I'm getting a bit frustrated myself.
*log pic*
76 failed attempts! GAH!!!!
farmpuma wrote:In the last twelve to fifteen hours I've tried four times (at an hour for each try) to upload this PoS WU! After an hour of sending data this server (which I'd like to smash with a ten pound sledge hammer) resets the connection and "tries another port!" And WTH is with this uncompressed upload file .. I mean really? .. a fourteen MB upload for a half MB download .. Really?
farmpuma wrote:For several weeks I've been seeing 171.64.122.139 trying to send me WUs *snip*. Uh Stanford, what the hell are these mystery WUs? *snip*
farmpuma wrote:... and an internet connection that can handle 50MB uploads. No SMP help for those of us on dial-up.
In other bad news - My 5903 GPU work unit with a 2606 at 71% on the secondary CPU core crashed at 42% completion and set idle for more than four hours during my sleep cycle. I will probably restart the 2606 at 50% when I have time to keep an eye on it, but for now the current 5903 has all of both the CPU cores.
Update to the update - I've now (17 March) lost a 5902 and a 5903 to EUEs. I may have forgotten to delete the work folder, queue.dat, and unitinfo.txt on the 5902, but I definitely deleted them before downloading the 5903. Maybe my outdated 174.88 driver has finally met it's match after over 700 work units or maybe these 590Xs aren't ready for prime-time, but since XP is still windowz I've tried a system restart. Currently at 38% and crunching on another 5903.
Ragnar Dan wrote:I keep slowing down my 9600GSO's shader clock, and it still get EUE's and UNSTABLE_MACHINE errors. I'm down to about a 6.5% overclock, and it still crashed on 5904 WU's. Very annoying. It has never performed this poorly before (right now showing < 1200 PPD). Quite a shame.
Ragnar Dan wrote:I keep slowing down my 9600GSO's shader clock, and it still get EUE's and UNSTABLE_MACHINE errors. I'm down to about a 6.5% overclock, and it still crashed on 5904 WU's. Very annoying. It has never performed this poorly before (right now showing < 1200 PPD). Quite a shame.
Ragnar Dan wrote:Well, I don't think I have any modern games that would push it. And I've yet to see any sort of artifacts or other evidence of failure on the screen. It's warm inside the case, certainly, and I may be able to cool it down some if I work on it long enough, but it's a big PITA and always causes annoyance getting inside there because of the sharp edges and the tenuously connected drive cables which almost invariably give me trouble when I put everything back together again.
Anyway, it seems to speed back up with some WU's, like most of the 768 point ones, so I think it's just the fact that most of these new WU's run more slowly on this 96-shader unit, and the new core FahCore_14.exe (got mine March 25) behaves in a rather Lewinsky-like manner, making more heat and producing little as a result.
Edit: farmpuma: I am running the Linux SMP client, so perhaps it's taking some of my output now, but it's a shame if they can't make it work like it used to.