Message boards :
Number crunching :
AK V8 ported release ap. issues, install, questions etc.
Message board moderation
Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · 8 · 9 . . . 19 · Next
Author | Message |
---|---|
Idefix Send message Joined: 7 Sep 99 Posts: 154 Credit: 482,193 RAC: 0 |
Hi, We will see soon SSE and MMX AK V8 also? See this posting of jason_gee: "SSE Yes. MMX ... 'Highly unlikely'" Regards, Carsten |
Gecko Send message Joined: 17 Nov 99 Posts: 454 Credit: 6,946,910 RAC: 47 |
Re: MMX Most of the current speed-up would not show in a MMX build. The amount of work required vs. minimal gain & the extremely small number of hosts (pre-Pentium III) that would benefit makes this impractical. The current 2.4V_Windows_x32_MMX application will likely remain the option for those w/ older pre-SSE CPUs. |
JDWhale Send message Joined: 6 Apr 99 Posts: 921 Credit: 21,935,817 RAC: 3 |
Hi, I know that I will not work on MMX client. A bit more work on SSE client today... nice improvements on Mid-AR WUs compared to KWSN_2.4V_SSE_MB client, though VHAR and VLAR WUs are bit slower :-( WU validation testing continues. That's all for today... time for bed. Regards, JDWhale |
_heinz Send message Joined: 25 Feb 05 Posts: 744 Credit: 5,539,270 RAC: 0 |
Hi Jason, hmmm, app shows the cores of the processor...not how many processors and all 8 cores are running at 100% full load, have a look at the wus... for your statistic data -------------------------------------------------------------------- Computer ID 4387433 Report deadline 12 Jun 2008 16:48:43 UTC CPU time 5654.313 stderr out <core_client_version>5.10.45</core_client_version> <![CDATA[ <stderr_txt> Windows optimized S@H Enhanced application by Alex Kan Version info: SSE4.1 (Intel, Core 2-optimized v8-nographics) V5.13 by Alex Kan SSE4.1 Win64 Build 42 , Ported by : Jason G, Raistmer, JDWhale CPUID: Intel(R) Xeon(R) CPU E5405 @ 2.00GHz Speed: 4 x 1995 MHz Cache: L1=64K L2=6144K Features: MMX SSE SSE2 SSE3 SSSE3 SSE4.1 Work Unit Info: ............... Credit multiplier is : 2.85 WU true angle range is : 0.406780 Flopcounter: 16458043843819.520000 Spike count: 1 Pulse count: 0 Triplet count: 0 Gaussian count: 1 called boinc_finish </stderr_txt> ]]> ------------------------------------------- regards heinz |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
... Ahh yes. This is a known issue resulting from using the code straight from 2.4V as a last minute add-on. Not to worry, it is only cosmetic. There are a few options I'm exploring and would like to use proper Intel/AMD recommended procedures to tidy this identification up. AMD Phenoms have a similar variation from expected, and don't show L3 cache either. Jason "Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions. |
_heinz Send message Joined: 25 Feb 05 Posts: 744 Credit: 5,539,270 RAC: 0 |
Hi Jason, Here is a very interesting compare of two V8 machines. Have a look and make your conclusion regards heinz |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
not too shabby Heinz!... On the surface it looks like maybe you got a free GigaHertz by running 64 bit sse4.1 :O Jason "Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions. |
_heinz Send message Joined: 25 Feb 05 Posts: 744 Credit: 5,539,270 RAC: 0 |
Hi, go through the results and found some -9 result overflows http://setiathome.berkeley.edu/result.php?resultid=850714783 http://setiathome.berkeley.edu/result.php?resultid=850712899 http://setiathome.berkeley.edu/result.php?resultid=850687649 http://setiathome.berkeley.edu/result.php?resultid=850687627 http://setiathome.berkeley.edu/result.php?resultid=850684125 http://setiathome.berkeley.edu/result.php?resultid=850683731 http://setiathome.berkeley.edu/result.php?resultid=850652348 http://setiathome.berkeley.edu/result.php?resultid=850569966 ------------------------------------------------------------- any ideas ? heinz |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
Hi Heinz, About three of those are waiting on wingmen for confirmation. The others are all verified mostly against stock as being noisy '-9 overflow' WUs. So all looks normal and nothing to worry about. One of those that is waiting for a wingman, has a compute error (but not from you) Jason [Edit: You'll see these quite a bit more often than on your P4 ... pretty good jump up in throughput there! ;D] "Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions. |
_heinz Send message Joined: 25 Feb 05 Posts: 744 Credit: 5,539,270 RAC: 0 |
Hi Jason, very nice, that there is no problem, looks like the machine runs great. Today its a little bit cooler, roomtemp 24 celsius, FB-Dimmm sensor shows 67 celsius, and cpu's around 45-54 celsius. So far I have the right cooling hardware choosen. In the next days we will see how the machine perform its rac. heinz |
Zule Send message Joined: 1 Jul 06 Posts: 52 Credit: 84,436,096 RAC: 0 |
I'm the guy that reads but never posts... But I'm so blown away by this upgrade that I have to tell you about it.. The box is a 5200 X2 on an MSI K9A2 Platinum board with 2 gig of Balistix mem at maxed mem speeds.. It has a very mild overclock.. I only pushed the bus from 200 to 202... Using the SSE3 V8 app and according to Seti's task list my sceconds have gone from 12-11,000 to 7-6,000 per WU.. I've scanned the results and so far no failures and the Spikes and whatever are the same of other boxes.... I've had a temp increase from 46c to 48c... So far 12 hours and rock solid... Please, Please, Please give me an SSE V8 app.. I have 6 Amd boxes that would LOVE to crunch it... I Like the KWSN 2.4 app, but this V8 is insane for my AMD box... Keep up the good work, and thank you so much for your efforts.. |
RandyC Send message Joined: 20 Oct 99 Posts: 714 Credit: 1,704,345 RAC: 0 |
That is indeed a very mild overclock: 0.5% I run a 4600 X2 on an MSI board too and I've got the FSB set to 210. That's a 5% overclock which I consider to be conservative. |
Zule Send message Joined: 1 Jul 06 Posts: 52 Credit: 84,436,096 RAC: 0 |
A baby step... Exhaustive testing.... Then another baby step... |
Sutaru Tsureku Send message Joined: 6 Apr 07 Posts: 7105 Credit: 147,663,825 RAC: 5 |
I have now with the new SSSE3x a: SETI@Home Informational message -9 result_overflow NOTE: The number of results detected exceeds the storage space allocated. wuid=264380332 And one wingman have stock app and well result.. An other Client error - Compute error I think it's not a problem with the new app.. It's the known issue that sometimes the app think it's an overflow, but it's not. Or? EDIT: The Client error - Compute error result is with stock or opt. app? |
[KWSN]John Galt 007 Send message Joined: 9 Nov 99 Posts: 2444 Credit: 25,086,197 RAC: 0 |
I would back off the OC on that machine. One of my Q6600s did that when I had it at 3.15. Backed it off to 2.7, and all is fine. The extra work (heat) that is being done will cause the -9s. Just FYI... Clk2HlpSetiCty:::PayIt4ward |
OG Send message Joined: 16 Mar 07 Posts: 2 Credit: 39,790 RAC: 0 |
The past day or so I seem to be getting some problems with granted credit. I usualy average around 50 credits for betewwn 5000-6000 seconds of work, but I'm showing several tasks where I've only recieved a couple of credits for that ammount of work. Ive tried reinstalling BOINC with no luck. 853911924 853850258 853006770 853053316 Theres about 12 in all off average. |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
The past day or so I seem to be getting some problems with granted credit. I usualy average around 50 credits for betewwn 5000-6000 seconds of work, but I'm showing several tasks where I've only recieved a couple of credits for that ammount of work. Ive tried reinstalling BOINC with no luck. Something not right there OG, you're claiming rather low there on those WUs... frankly that's one issue I've not seen before. I'll tuck this away in my brain looking for more information, but can't promise it's something I have an answer for it. Only thing that comes to mind is that inconsistency in reporting usually indicates an overclock pushed slightly too far... Jason "Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions. |
OG Send message Joined: 16 Mar 07 Posts: 2 Credit: 39,790 RAC: 0 |
Overclocking is a possability. I was pushing 3.2ghz on an X2 5000+ B.E. (16x200). Stability and temp wise it seems fine at that clock apart from these irratic credits. I've knocked it back to 2.8 to see if the credits return to normal. |
Karsten Vinding Send message Joined: 18 May 99 Posts: 239 Credit: 25,201,931 RAC: 11 |
OG: I have had exactly the same behaviour on my Athlon64 3800+, back when I pushed it to 2,5GHz+. The system would run perfectly stable for days on end, but would sometimes claim zero or very low credit credit, for validated results. On my PC, just lowering the clock 50MHz, or bumping the voltage a bit upwards removed the problem. My solution ended up being 50Mhz less, since bumping the voltage caused the fan on my CPU-cooler to accelerate to intolerable noise levels. |
Sutaru Tsureku Send message Joined: 6 Apr 07 Posts: 7105 Credit: 147,663,825 RAC: 5 |
EDIT: The Client error - Compute error result is with stock or opt. app? So one well and to now two errors.. So it was not an easy WU.. ;-) I don't think it's because of a bad OC.. I OC without changing the Vcore and the temps are fine.. ;-) Since months without probs.. and over 3 weeks with the new V8 and no problems.. Now the first problem.. I think it's the known issue with the stock app.. and the opt. app come from the stock app, so the opt. app have the bug also..?! Or not?? Opt. Team, please your meaning.. :-) For example, my V8 app have a well result but the stock app an overflow.. wuid=264426328 |
©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.