Cheating in RAC through Prime grid
First off, I would like to start off with how I came up to this situation and I do think it is possible to have been occurring for many of the people and I could not find any suggestion when I searched the net. So I was quite happily contributing in the asteroid at home even though it had many participant because even at full load my gpu temp (gtx 1050 ti) never went above 60 C and also that it used very nominal cpu time so that every threads could be fully utilized by the cpu project.
And there was a maintenance with the server of asteroid at home and no work unit to process. So, I tried changing the project to a few. First, I tried Amicable number, Einstein at home and also Collatz conjecture through the pool 1 GRC pool but I didn’t receive any gpu work units for a whole day. So, I changed to Prime grid where at least I got the work load and I was happy. But after a 1 hour of crunching I just found my mag to be like 0.2 which was like crazy high for a gtx 1050 ti. And I saw that most of the task I have had were giving computation error except for the shortest task of the PPSieve which took only 15 minutes to compute. I thought it was because the given numbers were not prime. So I just crunched for a few days just giving a 2-4 hour for prime grid and today I looked at the host page and I found that out of 124 work units assigned to me 99 showed error.
Here’s the links for my host computer for prime grid.
( http://www.primegrid.com//show_host_detail.php?hostid=908048)
And here’s the snapshot for the graph of my RAC for prime grid.
To say the least I was shocked that this was happening. The implication it made was tremendous. I was cheating the RAC even though it was unintentional. And I can be sure that many of other volunteers are also having the same condition. This also tells us that it is possible to cheat and get high RAC without getting much work done. Because BOINC cannot stop the user from aborting the larger work unit which give less credit. Which might affect the BOINC contribution.
All I want to do with this post is to raise awareness that this thing can happen and also the instance where it has happened. And I don’t know if it’s possible but can there be a check like if most of the work units are getting computation error then do something about it.
Also, if there is anything that can be done please reply.
Worthless post, very poorly researched and poorly written. I have checked your tasks, you haven't cheated anything - you received zero credits for those failed tasks which your faulty machine thrashed.
On the other hand, you managed to compute 24 valid CUDA PPS Sieve tasks and your RAC climbed a little because of that. PPS Sieve tasks are short but yield most credits (proportional to average runtimes) because you cannot discover a new prime number by crunching Sieve tasks, you are just doing 'preparation work' for LLR applications (which are more demanding and which can indeed uncover a new prime).
Posting poorly researched articles like this is usually the quickest way to get flagged. Join our Slack, we have a channel there called 'crunch-tinkering' (dealing only with BOINC crunching), you would get plenty of info there for a quality Steemit post.
SO if i keep crunching through prime grid with all these computation errors it will not be a peoblem right.
It is a problem, you are thrashing good tasks which have to be resent then, wasting resources. In your Primegrid account, choose to receive only PPS Sieve tasks and you'll be fine.
And change the title of this post, if it's not too much inconvenience, because cheating is an ugly word, usually associated with FUD and everything is working as it should in this case.
You should look for answers / ask on PrimeGrid forums. It's better not to get errors.
Well first of all 0,2mag for gtx 1050 ti isnt crazy high. Secondly did you check these false tasks? When you check them you can see from tasks with error you get from 0 to few hundred credits, when for completed and vaildated tasks you get over 3000 points.
You've done 24 tasks successfully within 2 days or so, each rewarding you with 3,371 credits. It's natural your RAC is already up to 6,500. Looks OK.
no i mean this much of task jsttok atual 15 * 24 minutes only.
You mean you gained a lot of credit for only 15*24 minutes of computation? If so. Please note that not all projects grant credit in an equal way.
Do some research on the topic. There is lots of info in #gridcoin. I myself also have been publishing about it lately.
I don't understand how you are cheating the RAC? Did you notice you also got validated tasks with big credit numbers that accumulated to the end RAC.
I bet if I run my experiment on them my calculations just turn out fine to the numbers you see. In fact. I will do it tomorrow and make a post about it.
Are those 124 tasks all tasks since you began?
As promised: https://steemit.com/gridcoin/@jefpatat/debunking-rac-fud
I didn't read all, but to this:
Only the PrimeGrid admins can influence/correct the RAC points assigned (or do something about cheating, ...): contact them in their forum or via PM
Congratulations @buggy143! You have completed some achievement on Steemit and have been rewarded with new badge(s) :
Click on any badge to view your own Board of Honor on SteemitBoard.
For more information about SteemitBoard, click here
If you no longer want to receive notifications, reply to this comment with the word
STOPIn this lobby, will there be T.P.? For my bunghole?
Congratulations @buggy143! You have received a personal award!
Click on the badge to view your own Board of Honor on SteemitBoard.
Congratulations @buggy143! You received a personal award!
You can view your badges on your Steem Board and compare to others on the Steem Ranking
Vote for @Steemitboard as a witness to get one more award and increased upvotes!