Message boards : Number crunching : How To Captivate And Retain Power Crunchers
Author | Message |
---|---|
XS_The_Machine Send message Joined: 2 Jan 06 Posts: 47 Credit: 4,612,053 RAC: 0 |
I think it may be benificial to start a thred and have the power crunchers share their opinons as to what they expect out of a project. This in no way puts down the average users as most say they don't care so much for points as long as they are fair across the board and the others who install and forget.Also they would be welcomed to share if they so desire. This may help project managers who will have to deal with such issues in the future. Sorry for the off topic. Just trying to help in a small way. Delete as necessary i won't take it personally :D |
Mod.Tymbrimi Volunteer moderator Send message Joined: 22 Aug 06 Posts: 148 Credit: 153 RAC: 0 |
|
Vester Send message Joined: 2 Nov 05 Posts: 258 Credit: 3,651,260 RAC: 521 |
[*]Some power crunchers are not members of any team, but they take part in teams' discussions. [*]On a friendly message board which allows members to post and become aquainted, teams will form. Such teams are loyal to the project. [*]Some members may arrive with one computer, become addicted to crunching, and add computers. One recently departed member has about 30 computers in his residence. [*]Allow members to run Rosetta@home without having to run BOINC. [*]Develop a client with a queue server that can be run on members' networks without having to be run stand-alone. [*]Stop deleting posts. Let them post it, sober up, and have to live with what they said. [*]Implement personal mail. [*]Add a preview feature. Chances are that this message board won't display my post correctly. |
Ananas Send message Joined: 1 Jan 06 Posts: 232 Credit: 752,471 RAC: 0 |
This one is in the features list of the latest Pirates improvements, maybe they would be willing to share. They added more interesting things too like attachments, a better quoting system ... Their forum style isn't the best as the quotes stick out more than the new replies but that could sure be fixed in the CSS file. |
dcdc Send message Joined: 3 Nov 05 Posts: 1832 Credit: 119,675,695 RAC: 11,002 |
I think that would be a big help. Also, links to IRC/Messenger etc if people are willing to make them available, so people can communicate more easliy. |
Jayargh Send message Joined: 8 Oct 05 Posts: 23 Credit: 43,726 RAC: 0 |
Please define the term "Power Cruncher"...At what point do I go from being a user to elite? 2000 rac? 5k 10k? What is YOUR definition? I am not one in my definition but am curious especially from the original poster but also from anyone who can quantify this term I hear used over and over on many boards but am really unclear where the cut-off is.Is it the top 100 users of a project? Top 50? Or is it rac only? I really don't understand this term as if you were to give a value and I was right below that value now I'm not a Power Cruncher? Is my team too low in rac to have a personal visit by Dr Baker in our forum if we threatened to pull out? What is the purpose of this post? To Captivate and retain exactly who? I noticed all respondees have much less credit than the original poster so will this thread improve the project since most are probably not Power Crunchers.? |
XS_The_Machine Send message Joined: 2 Jan 06 Posts: 47 Credit: 4,612,053 RAC: 0 |
I had never used the term power cruncher until that post it was a term i have seen used in other post so i used it as a general term. My thoughts of what a power cruncher is a person who uses multiple boxes devoted to a single project who tunes these rigs for maximum output. In general i see comments like the top users get board with a project after they reach the top position so i thought it may be of some benefit if some of these top users would list some things that may keep them captivated. In my observations as a member of such a team here are a few that i see come up in discussions. You must have a reliable stats server with backup takes the #1 spot. (we are stats junkies and need several fixes daily) also this is how we keep track on the performance of individual boxes. Projects that waste cpu power & electricity with quorums of 2-3 are not usually chosen. The project has to be worthy of all this expense in time and money. And i have personally have never seen mention that a project developer should visit our forum. It's nice but it can cause problems. I hope everybody has moved beyond that because it was an honest mistake made by someone who admittedily has very little experience in forums. Also these people need to keep there thoughts on science not a bunch of hard to please crunchers:) All comments made are personal and i in no way represent the XtremeSystems team. XS The Machine is a sub team with a membership that varies at any given time, i personally ran 2 opteron 170 at 2800 watercooled 24/7 on gentoo 64bit because before 5.5.0. these were the top producing rigs with a rac of 2200. I also ran 2 3.4EE on 5.5.0. sometimes 24/7 sometimes not due to heat with 4 rigs running in the same room. I was just a average user with team XS. heater |
kevint Send message Joined: 8 Oct 05 Posts: 84 Credit: 2,530,451 RAC: 0 |
Well I am not sure if what I say will make any difference or not but I will try. A power cruncher is a RAC/Credit junkie in the simplest form. The project must be worth crunching, something of interest, and worthwhile. Credit for power consumption must be considered. It is one thing to run 1, 2, 3 or even 5 dedicated cruncers 24/7 but power crunchers seem to have more. 15+ (I have well over 30 last count I had several crash and burn on me recently, most are just dedicated crunching machines) These machines for most part run in my house. All are personal, and I pay the power bill. A power cruncher wants equal credit for $$$ spent, both on equipment and on credits. Currently this is not working, with cost of power going up, it seems that project devs for what ever reason have started to restrict the credit granting. It quickly becomes not worth the effort to support a farm of machines crunching 24/7 when projects are stingy with credits. It seems (no flame bait intended) that those that are running 1 or 2 or maybe 3 machines, or are running them from work etc.. and are not responsible for the power and upkeep of these machines cry and whine about too much credit being granted. Grant credit where credit is due, credit is not worth anything, can not be spent but is a way that projects can reward those that crunch. Compitition is a powerfull motivator and when projects become stingy with credits, power crunchers will move on to projects that pay them more for the money they have spent on equipment, power, and upkeep of these farms. $8.00 (or credits) per hour is not enough IMO - it barley pays the power bill. SETI.USA |
Tony DeBari Send message Joined: 13 Apr 06 Posts: 12 Credit: 2,944,090 RAC: 0 |
I ask the following not to be confrontational but out of genuine curiosity: Why do, as Kevin says, PowerCrunchers believe they are entitled to BOINC credits for things like the time and money they spend acquiring, upgrading, tweaking, and running their rigs? Are we not all volunteers donating our resources - however vast or meager they may be - to the project? If that is true, then surely the only "compensation" to which we are entitled is for the science we actually contribute, and perhaps an occasional "Thank You" from the project staff and recognition for a significant result or achievement . I understand that competition is a (the?) big motivator for the PowerCrunchers, and it is through the competitive aspects of BOINC that they bring to bear massive amounts of computing power and produce vast amounts of science output for this (or any) project. But at what point does the paradigm change from donating resources to the project, to investing time/money/etc. in the project and expecting a return on that investment in terms of (for lack of a better analogy) "interest" payed in BOINC credit above and beyond that which measures the amount of science one produces? Perhaps, as a generally non-competitive person and self-proclaimed ZeroRACer, I will never understand what drives the PowerCrunchers to do what they do. But I believe that it is this difference of opinion as to what BOINC credits are and how they should be calculated that creates the disconnect that we saw come front and center recently, and continues to reverberate throughout this forum. Regards, -- Tony |
zombie67 [MM] Send message Joined: 11 Feb 06 Posts: 316 Credit: 6,621,003 RAC: 0 |
I ask the following not to be confrontational but out of genuine curiosity: I understand the question, and I understand your not understanding. But you're missing the point. This is not a thread about why. This is a thread about how. Let's get back on topic. You want me? Give me better detailed stats, especially machine-level detail, over time. We need a way to see if one of our machines has fallen down real-time. And graphs. And get rid of RAC. No one understands it really. Find a better measure. Reno, NV Team: SETI.USA |
Whl. Send message Joined: 29 Dec 05 Posts: 203 Credit: 275,802 RAC: 0 |
Crunching is a very boring thing to do. Keep it interesting apart from the science. Which it was kinda before the blandness of the new credit system. Oh, and one very important thing is, dont insult the members. |
Vester Send message Joined: 2 Nov 05 Posts: 258 Credit: 3,651,260 RAC: 521 |
@zombie67: If we understood why people crunch, then we could better understand how. Here is one of my favorite articles about distributed computing: http://www.onlamp.com/pub/a/onlamp/2003/05/22/distributed.html. Distributed Computing: Distributed Communities Continued at ONLamp... |
Leviathan18 Send message Joined: 18 Mar 06 Posts: 8 Credit: 156,163 RAC: 0 |
im not producing here anymore but im going to leave my comment in this matter. for me is easy, if i run a DC project i will run it like a company, i need pc so i look for people to crunch for me for that i give them credits i have to show my project is time worthing like this one i wont attract people with a project to find how many sand grains are in the planet so i make a science project looking for cures of major problems like cancer aids etc, then i threat my comunity like people in a hotel 5 stars hotel treats guests like kings and if they have more computer power like gods (imagine someone at the presidential room) if the top is complaining because no one takes care of them you should see what you have to do to keep them happy they are the power moving your project, the money running the hotel if you dont like to threat people just dont make this a DC project |
Sequest Send message Joined: 5 Jun 06 Posts: 1 Credit: 672,228 RAC: 0 |
My only real complaint about the new credit system is that I can never catch anyone. The total credits are so high relative to the new reduced RAC that all I can do is watch everyones stats sit unchanged or changing glacially. Where I used to be able to catch another power crucher in a few days/weeks if I worked my crunch farm hard, now its months. Its boring. I dont expect a big wet kiss from R@H for my power bill, but I expect to be challenged or entertained.<p><p> Sure 'its all in the name of science' and stuff, but lets face it that if all projects completely removed credit and competition then NO ONE would participate. Fighting the good fight is great and all, but it leaves me kinda ho-hum. Im starting to shut down some systems at night, dont plug the laptop in overnight, not worrying about crunching on them. I used to run two laptops in my car during commuting for a few extra credits when I was chasing down an adversary but now whats the point. I think there is a big difference between being .000000001% of the total grid, and needing just 100 credits/day to get in the top 100.<p><p> Someone suggested starting a new project, all scores reset to zero. Im all for that. FWIW I run ~24 crunchers. Dont know if that makes me a power user or not. |
tralala Send message Joined: 8 Apr 06 Posts: 376 Credit: 581,806 RAC: 0 |
My only real complaint about the new credit system is that I can never catch anyone. The total credits are so high relative to the new reduced RAC that all I can do is watch everyones stats sit unchanged or changing glacially. Where I used to be able to catch another power crucher in a few days/weeks if I worked my crunch farm hard, now its months. Its boring. I dont expect a big wet kiss from R@H for my power bill, but I expect to be challenged or entertained.<p><p> Well that is a dilemma. It is indeed demotivating when the credit system changes and you get about a third of your former credit (in your case it looks more like 40%). That cements the current standings and makes it much harder to overtake people. However resetting the stats will annoy other users as was expressed several times. What do we learn? Never let the credit system for too long slip in an unbalanced situation since you can't get out of that situation without alienating some people. Btw. upward adjustment of the credit system would have caused great discomfort with other projects and would have also resulted in a loss of members (besides the fact that is against the BOINC spirit). |
Vester Send message Joined: 2 Nov 05 Posts: 258 Credit: 3,651,260 RAC: 521 |
You passed 19 yesterday, Sequest. I certainly won't catch you! http://img321.imageshack.us/my.php?image=capture31082006172707zg3.jpg |
rochester new york Send message Joined: 2 Jul 06 Posts: 2842 Credit: 2,020,043 RAC: 0 |
e bay has different color stars for their milestones something like that could be done here for rosetta |
dcdc Send message Joined: 3 Nov 05 Posts: 1832 Credit: 119,675,695 RAC: 11,002 |
My only real complaint about the new credit system is that I can never catch anyone. The total credits are so high relative to the new reduced RAC that all I can do is watch everyones stats sit unchanged or changing glacially. Where I used to be able to catch another power crucher in a few days/weeks if I worked my crunch farm hard, now its months. Its boring. I dont expect a big wet kiss from R@H for my power bill, but I expect to be challenged or entertained.<p><p> It's much more difficult to catch those above you now than it was previously if the gaps are increased due to the old system, but I think the competition should be much better now - the stats are accurate and you get credit for the CPU power you provide rather than for an arbitrary benchmark! My lappy (1.86GHz P-M Dothan) is getting around 20% more than it's claimed credits now. I didn't and couldn't know it was slightly outproducing my 2GHz Athlon 64 (Venice) even though the venice is on 24/7 and the lappy isn't quite. I think it's the psychological issue with the smaller numbers coming through for those that were using optimised clients. As tralala mentioned, credit needs more emphasis in new projects to aviod problems later. P.S. I think a RAC of 6k qualifies you as a power cruncher! |
Message boards :
Number crunching :
How To Captivate And Retain Power Crunchers
©2024 University of Washington
https://www.bakerlab.org