Message boards :
Number crunching :
again less credits?
Message board moderation
Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · Next
Author | Message |
---|---|
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
Is this why my RAC has dropped from 72,000 back in May to 61,000 today? Or is there some other sinister workings going on? It's all a sinister plot. SETI@Home isn't based out of the Space Sciences Lab, it's really operated out of the Art department. It's a "performance art" piece inspired by the work of J.S.B. Boggs. First they convinced us that credits had intrinsic value, and now they're devaluing the currency to see how we'll all react. Google J.S.B. Boggs if you want to know more. |
-=SuperG=- Send message Joined: 3 Apr 99 Posts: 63 Credit: 89,161,651 RAC: 23 |
Hah... glad there is still humour in here.. Art Department... That explains everything.. Boinc Wiki "Great spirits have always encountered violent opposition from mediocre minds." -Albert Einstein |
Labbie Send message Joined: 19 Jun 06 Posts: 4083 Credit: 5,930,102 RAC: 0 |
Is this why my RAC has dropped from 72,000 back in May to 61,000 today? Or is there some other sinister workings going on? Nah, it's run out of the psychology department, they're studying mob mentality. Calm Chaos Forum...Join Calm Chaos Now |
Brutus Send message Joined: 26 Apr 00 Posts: 3 Credit: 6,871,910 RAC: 0 |
You're absolutely right... Looking at my pending history of Astropulse units the oldest has a pending value of 1267 and there is a constant decline in value over 30 or 40 units with the most recent being 1225. There seems to be a similar trend in the MB units but it's harder to tell because there seems to be more variation in the unit values. You know I started running Seti in early 2000 and ran for 4 years straight before getting side tracked. It was fun. There were a number of 3rd party applications that showed you star maps etc. The site provided some detail on the science collected, potential candidate signals, etc. It wasn't huge but it was something simple that gave back to the users running the app. I came back 2 months ago when I saw something about a CUDA application. I've been running the app for 2 months on a couple of computers and picked up 1.1M points. I'm nearly ready to uninstall the application again. There is little on the site that gives back the users. There's almost no status on the science, there's no pretty star graphs, no sky maps showing where we've looked etc. I realize this is all volunteer but this exceptionally demotivating. I guess we're supposed to run this application on our computers, investing time and money and get virtually no feed back. There are 6 - 7 month time gaps in the science news letters which are superficial at best. The only satisfaction is to see some meaningless score rack up. I'd like a bit more than that. The stats are fine but the amusement is limited. I know I'm probably going to piss a bunch of people off by making these comments - not my intent. |
zoom3+1=4 Send message Joined: 30 Nov 03 Posts: 66359 Credit: 55,293,173 RAC: 49 |
Is this why my RAC has dropped from 72,000 back in May to 61,000 today? Or is there some other sinister workings going on? Are You sure It isn't a collaboration between the Art Department and the Psychology Department? ;) Savoir-Faire is everywhere! The T1 Trust, T1 Class 4-4-4-4 #5550, America's First HST |
W-K 666 Send message Joined: 18 May 99 Posts: 19407 Credit: 40,757,560 RAC: 67 |
You're absolutely right... The sky map for the tasks you complete is still around but you also needs a program to log the details. See BOINC add on S/ware, BoincLogX and Mapview. |
Mumps [MM] Send message Joined: 11 Feb 08 Posts: 4454 Credit: 100,893,853 RAC: 30 |
The sky map for the tasks you complete is still around but you also needs a program to log the details. See BOINC add on S/ware, BoincLogX and Mapview. I didn't think either of those apps worked with AstroPulse WU's. Am I mis-remembering? |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
Is this why my RAC has dropped from 72,000 back in May to 61,000 today? Or is there some other sinister workings going on? Ah, but isn't Art about psychology -- how people perceive and interact with the environment and others in it? Boggs is interesting because he "draws" money, and then tries to spend it. When people agree (or don't agree) to a transaction, he asks them what makes his "art" different from the government-sponsored "art" of a $10 bill. ... and it's the same thing here, except they aren't going to any extra trouble to actually create the "art" we'd normally call money. It's a whole economy of cobblestones without actually spending a dime. |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
On threads like this, we often see words like "sinister" whenever credit is adjusted. Invariably, someone writes about what the project is "doing to" those of us whose computers actually do the work. If you think this is done out of malice, look at it from the other side: Picture this from Eric Korpela's point of view. He's got a lot invested in SETI@Home being successful. It doesn't make sense for someone with a lot invested in the project to drive people away. It follows that one of his goals is fair and accurate credit. We can argue about the implementation of that goal, but when we start questioning motives, sorry, but that does not make any sense at all. |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
The sky map for the tasks you complete is still around but you also needs a program to log the details. See BOINC add on S/ware, BoincLogX and Mapview. Your memory is fine. The logging of information needed for Astropulse display on the map does not work, apparently there is code for the old 2004/2005 Beta but the current result format is different. OTOH, BoincLogX will keep track of Astropulse work if you want a record of which WUs you've done and how long they took, etc. Joe |
-=SuperG=- Send message Joined: 3 Apr 99 Posts: 63 Credit: 89,161,651 RAC: 23 |
I wholeheartedly agree. So who is questioning motives? On threads like this, we often see words like "sinister" whenever credit is adjusted. Invariably, someone writes about what the project is "doing to" those of us whose computers actually do the work. Boinc Wiki "Great spirits have always encountered violent opposition from mediocre minds." -Albert Einstein |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
I wholeheartedly agree. So who is questioning motives? In this post. it says: Is this why my RAC has dropped from 72,000 back in May to 61,000 today? Or is there some other sinister workings going on? Following posts go on to question the motives behind these sinister workings. ... and invariably, the thread works its' way around to how we volunteers are being cheated by declining credit -- in a world where a million BOINC cobblestones won't buy a cup of Coffee at the local McCafe, let alone Starbucks. |
perryjay Send message Joined: 20 Aug 02 Posts: 3377 Credit: 20,676,751 RAC: 0 |
I bet it's all those heavy hitters shaving points off us little people!! :) PROUD MEMBER OF Team Starfire World BOINC |
-=SuperG=- Send message Joined: 3 Apr 99 Posts: 63 Credit: 89,161,651 RAC: 23 |
I bet it's all those heavy hitters shaving points off us little people!! :) Heh.. quite possible perryjay, quite possible... har har. Boinc Wiki "Great spirits have always encountered violent opposition from mediocre minds." -Albert Einstein |
Chelski Send message Joined: 3 Jan 00 Posts: 121 Credit: 8,979,050 RAC: 0 |
There are 2 main points I see. 1. Equal work gets equal credit (at any given time). If anyone who runs a MB get the same amoount of credit then this is fine. Currently it is more or less true for all CPU work (barring any unfortunate pairing with old clients). There's discrepancy between MB-CPU vs MB-CUDA in the sense that an exclusive CUDA pairing get slightly higher credit. The MB and AP discrepancy will presumably get fixed once the stock AP performs better compared to the optimised one. 2. Equal work gets equal credit over time This is the subject of this thread. I don't think it is a big issue, because the real world work in the same way that a unit of production is worth less over time and the only way business can survive is to exceed the price erosion of manufactured goods with increase in productivity and material efficiency. Many years ago, selling a single core x86 CPU taking 40 hours to do a classic work, in a package <500 pins earns Intel more money than what they sell with 4 cores, 8 threads with >1000 pins or those 2 cores multi chip packages. The problem only come in because the project credit granting mechanism is adjusting down towards a hopefully steady state convergent point despite stated target that credits per hour should theoretically stay constant over time. Personally this is not a big issue as it affects everyone the same and that it is adjusting down (in line with real market), and in the reassuring finality that credits are essentially worthless outside the confines of personal interpretation. |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
The problem only come in because the project credit granting mechanism is adjusting down towards a hopefully steady state convergent point despite stated target that credits per hour should theoretically stay constant over time. The project is not trying to converge on constant credits over time. Originally, BOINC granted credit based on the benchmark * time. This follows from the definition of a cobblestone, which is 1/100th of a day worth of work from a machine with specific benchmark scores. But SETI doesn't use benchmark * time to calculate credit. It's based on counting FLOPs and an internal conversion faction which was figured by hand, and is a bit off. ... and hard to correct. So they added a script to tune the multiplier and bring credit into line with the original standard. Eric Korpela described the reason, and the process, in this post. The whole thread is worth reading. [edit]Just keep in mind that as with so many threads, there is as much misinformation and conjecture as there is fact.[/edit] The goal is not to keep credits-per-hour constant, but to keep credits comparable to the mythical 100 cobblestone computer. |
W-K 666 Send message Joined: 18 May 99 Posts: 19407 Credit: 40,757,560 RAC: 67 |
The problem only come in because the project credit granting mechanism is adjusting down towards a hopefully steady state convergent point despite stated target that credits per hour should theoretically stay constant over time. One thing to remember when reading Eric's post and working out what his script does. It was before gpu crunching came along, so unless the script is changed to exclude gpu tasks, you might be able to work out what the effect of incorrect time reporting and over claiming is. |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
It is entirely possible that CUDA has moved the median. The reason for using the median is to remove the "outliers" -- machines that are unreasonably slow (or really old versions that claim "0" credits) and machines that are amazingly fast. What we don't know is: are there enough CUDA results that the median machine is crunching on a GPU and not a CPU. By which I mean, I don't know. What I was talking about in my original post is "what is the goal" because we keep seeing people say "constant credits per hour" and that credit will go down as machines get faster -- and that is not the intent. Your question goes to "does the implementation do what it is supposed to do?" which is an important question, but it's a different question. |
Dirk Villarreal Wittich Send message Joined: 25 Apr 00 Posts: 2098 Credit: 434,834 RAC: 0 |
Hi folks! I have realized that since the 7th of June, the BOINC Statistics of the World is showing zero credits granted on my account. Any ideas why? My PC has been crunching all that time as usual. I´ve never seen this behavior before. Some kind of readjustments? Crisis has reached the BOINC Credit Bank account as well!!! |
Leopoldo Send message Joined: 4 Aug 99 Posts: 102 Credit: 3,051,091 RAC: 0 |
Hello, Dirk! ...BOINC Statistics of the World is showing zero credits granted on my account. Did You mention about Synergy's "BOINC Statistics for the WORLD!"? 2 messages here By speaking shortly, there is red text at Synergy's main page: Due to server issues, SETI's Stats will no longer be included in Synergy's Updates. |
©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.