Message boards : Number crunching : Why do some WU give 59,000 credit and some 70,800 for same run time?
| Author | Message |
|---|---|
|
I can't seem to find a pattern. Moreover, the top host are getting 70,800 credits every time. | |
ID: 29632 | Rating: 0 | rate:
 /   Reply
Quote
| |
|
Bedrich Hajek Send message Joined: 28 Mar 09 Posts: 486 Credit: 11,386,082,230 RAC: 8,982,437 Level  Scientific publications      | |
|
You get a 50% bonus for getting the unit complete and valid within 24 hours, and 25% bonus for getting it within 48 hours. | |
| ID: 29633 | Rating: 0 | rate:
/ Reply
Quote
| |
|
Oh cool, thanks. Math doesn't quite add up but close enough, didn't know that. | |
| ID: 29634 | Rating: 0 | rate:
/ Reply
Quote
| |
 skgivenVolunteer moderator Volunteer tester  Send message Joined: 23 Apr 09 Posts: 3968 Credit: 1,995,359,260 RAC: 0 Level  Scientific publications  | |
|
Going by your runtimes, your GTX560M should be able to finish most tasks within a day, so your cache is too high. If you change it to something like 0.01 and 0.01 then you should be able to download, run and upload tasks within 24h. If your cache is 0.50/0.50 then tasks could wait for up to a day before they even start running and you would definitely miss the 50% bonus. | |
| ID: 29635 | Rating: 0 | rate:
/ Reply
Quote
| |
|
ExtraTerrestrial Apes Volunteer moderator Volunteer tester  Send message Joined: 17 Aug 08 Posts: 2705 Credit: 1,311,122,549 RAC: 0 Level  Scientific publications | |
|
Side note: GPU-Grid is more powerful for the researchers the quicker they get results back (since the next WU depends on the results of a previous one). Hence the bonus credits for quick returns, which is unusual in the BOINC world. | |
| ID: 29669 | Rating: 0 | rate:
/ Reply
Quote
| |
|
Jozef J Send message Joined: 7 Jun 12 Posts: 112 Credit: 1,140,895,172 RAC: 17,863 Level Scientific publications | |
|
GPU-Grid is more powerful for the researchers the quicker...lol you must joke about this..You have a write-we still we have a problem of Kepler and 4.2 cuda ms.noelia works... Therefore we do not support GPUGRID counting on Titan nvidia cards. And on the next 6-9 months, we just expect Maxwell architecture-support is far far in heaven. So we are already two architecture behind ... only sarcasm, god day FOR ALL | |
| ID: 29677 | Rating: 0 | rate:
/ Reply
Quote
| |
|
mikey Send message Joined: 2 Jan 09 Posts: 298 Credit: 6,653,553,182 RAC: 15,004,121 Level  Scientific publications  | |
GPU-Grid is more powerful for the researchers the quicker...lol you must joke about this..You have a write-we still we have a problem of Kepler and 4.2 cuda ms.noelia works... Therefore we do not support GPUGRID counting on Titan nvidia cards. And on the next 6-9 months, we just expect Maxwell architecture-support is far far in heaven. So we are already two architecture behind ... only sarcasm, god day FOR ALL To be quite honest it does NOT take the latest technology to still do usable research, lots of researchers, and businesses, are STILL using Windows XP yet they all get their work done just fine. YES absolutely the latest and greatest hardware is faster, but if the old stuff works and the new stuff requires alot of time and energy that is always at a premium, sometimes it is just better to keep on keeping on. I used to work for one of the richest County Governments on the East Coast of the US, a 100 billion dollar a year budget, yet they are STILL running 15 year old software because it still does the job. Sure they could buy new stuff and YES absolutely it is preventing some hardware from being upgraded, but it is working as intended. Now lately the IT folks have started to complain about the cost of maintaining the old hardware, and that IS driving the look for a newer software developer, but that is still a few years away at best.  | |
| ID: 29691 | Rating: 0 | rate:
/ Reply
Quote
| |
|
ExtraTerrestrial Apes Volunteer moderator Volunteer tester Send message Joined: 17 Aug 08 Posts: 2705 Credit: 1,311,122,549 RAC: 0 Level Scientific publications | |
GPU-Grid is more powerful for the researchers the quicker...lol you must joke about this..You have a write-we still we have a problem of Kepler and 4.2 cuda ms.noelia works... Therefore we do not support GPUGRID counting on Titan nvidia cards. And on the next 6-9 months, we just expect Maxwell architecture-support is far far in heaven. So we are already two architecture behind ... only sarcasm, god day FOR ALL No, I'm not. And problems with new developments (new features in the CUDA 4.2 app used by Noelia, Titan) don't change at all how the existing code works. Believe me, the team is hard at work behind the scenes. They've achieved great things in the past and I'm sure they'll also get around these problems.. it's just that noone knows yet how long this will take. MrS ____________ Scanning for our furry friends since Jan 2002 | |
| ID: 29716 | Rating: 0 | rate:
/ Reply
Quote
| |
|
Great observation, great suggestion, thanks a bunch! | |
| ID: 29748 | Rating: 0 | rate:
/ Reply
Quote
| |
Message boards : Number crunching : Why do some WU give 59,000 credit and some 70,800 for same run time?
