Message boards :
Number crunching :
I'm falling, I bought a parachute. From 100% AP, to 100% MB.
Message board moderation
Previous · 1 . . . 4 · 5 · 6 · 7
Author | Message |
---|---|
juan BFP Send message Joined: 16 Mar 07 Posts: 9786 Credit: 572,710,851 RAC: 3,799 |
Any resemblance is not purely coincidental: https://www.youtube.com/watch?v=dRB-woVjlFY hope you have a soft place to land too. |
James Sotherden Send message Joined: 16 May 99 Posts: 10436 Credit: 110,373,059 RAC: 54 |
Sten, Are you running lunatics on your computer? Sten I have set no new tasks on the I7 920. It will take a few days to clear out the work. And then uninstall Lunatics. Would you prefer I open my own test thread, Or can I post on your thread? [/quote] Old James |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
Just to save some time on testing I would stress that GPU stock and opt are quite close by performance, especially if stock app (it's slightly outdated opt GPU app actually) will be tuned a little with more optimal for fast GPU params. So one could infer results of such testing perhaps. On CPU side we have bigger discrepance in performance, but using stock AP I would call torture and torture forbidden by international law :DDD SETI apps news We're not gonna fight them. We're gonna transcend them. |
James Sotherden Send message Joined: 16 May 99 Posts: 10436 Credit: 110,373,059 RAC: 54 |
I think for the sake of the sanity of the participants on the board (that all knows what we're talking about) it would be better if you have your own thread for the stock test :-) Go for it James, that test should be interesting indeed. Ok then, I will start my own thread. That way folks wont get confused of who is doing what. I am going to torture myself and do MB and AP on the CPU, Plus run MB on my GTS-250. I want to see if there is any RAC advantage to run without op apps. But I might try running just MB straight up after I see what happens. The hard part will be getting an average RAC for this machine. The stats graph looks like mountains and valleys. You can definatley see when we run out of AP work. I just might take the high and low and do the average. So a few more days untill I clear my cache, And then I will open my own thread. Sten, Thanks for doing this test. You have inspired me to do my own. I still say you will level out at 8,000. [/quote] Old James |
Gone Send message Joined: 31 May 99 Posts: 150 Credit: 125,779,206 RAC: 0 |
Sten almost hits skydiver on way down. http://youtu.be/gzYJaQ0h4Dg Here we see Sten disguised as a meteorite flashing passed a normal skydiver..... Reg |
Wiggo Send message Joined: 24 Jan 00 Posts: 34744 Credit: 261,360,520 RAC: 489 |
I've been doing a test myself Sten with this old setup that I have a thread on (and just updated, so I'll just quote my latest update here if you don't mind). Update. Keep up the good work. Cheers. |
HAL9000 Send message Joined: 11 Sep 99 Posts: 6534 Credit: 196,805,888 RAC: 57 |
I think for the sake of the sanity of the participants on the board (that all knows what we're talking about) it would be better if you have your own thread for the stock test :-) Go for it James, that test should be interesting indeed. I pull the numbers from statistics_setiathome.berkeley.edu.xml using a .bat and average them to get my average RAC. So it spits out "Host: 5837483 Average: 8030 for 58 days" SETI@home classic workunits: 93,865 CPU time: 863,447 hours Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[ |
James Sotherden Send message Joined: 16 May 99 Posts: 10436 Credit: 110,373,059 RAC: 54 |
I will take your advice and run 100% MB. Then when stable try 100% AP, running just stock apps. By the looks of my cache I still have at most two days to go. [/quote] Old James |
rcthardcore Send message Joined: 23 Nov 08 Posts: 48 Credit: 1,306,006 RAC: 0 |
I see what you mean by CreditScrew. Definitely feeling it here. |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
I will take your advice and run 100% MB. Then when stable try 100% AP, running just stock apps. I'm not sure though what this experiment should prove or disprove. Lets outline this to make it more interesting. It's said, that CreditScrew should pay equally for equal work done. Also it's said that the "master copy" is stock app. Also we know that looks like only CPU stock app involved, GPU gets screwed credits always no matter opt or stock. So, if at least some of postulates of CreditScrew are fulfilled one should not see difference in average RAC running stock CPU MB and stock CPU AP. Looks like you want to test this statement/prediction. Well, Eric tested this on beta (and even on main?) server-side AFAIK, but case-study on particular host will be interesting too. So, prediction is - no change in RAC. Will see ? :) SETI apps news We're not gonna fight them. We're gonna transcend them. |
William Send message Joined: 14 Feb 13 Posts: 2037 Credit: 17,689,662 RAC: 0 |
scaling. best stock MB is AVX, best stock AP is what, SSE2? higher SIMD, less credit, if you want it put simply. but do by all means run an experiment. Data is always good. The implementation of CreditNew does not match the whitepaper. over and out A person who won't read has no advantage over one who can't read. (Mark Twain) |
James Sotherden Send message Joined: 16 May 99 Posts: 10436 Credit: 110,373,059 RAC: 54 |
You all raise some great points. Im heading to work now, But I will prepare what I want to do and maybe try it in 3 steps. When I get home tonight. @ William- Would you say Stens idea about just renaming the App_ config.xml file to another name untill after my tests are done is a good workaround? [/quote] Old James |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
scaling. best stock MB is AVX, best stock AP is what, SSE2? Stock CPU AP has no SIMD at all. Joe |
©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.