PrimeGrid
Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants

Community

Leader Boards

Results

Other

drummers-lowrise
1) Message boards : Number crunching : Nvidia 500 series vs 600 series (Message 45974)
Posted 3551 days ago by SiKTheGreatOne
I am going to assume the specifications listed on this website are accurate.

http://en.wikipedia.org/wiki/Comparison_of_Nvidia_graphics_processing_units

The video card in the 500 series that looked most appealing to me was the GTX 560 Ti since it had the highest GFLOPS/Watt ratio which was 7.43. However, the GTX 650 has a GFLOPS/Watt ratio of 43.88.

GTX 560 Ti: 1263.4 GLOPS with 170W TDP (PCIe 2.0)
GTX 650: 3072 GFLOPS with 70W TDP (PCIe 3.0)

Would it be wrong to assume that the GTX 650 would be a lot better based just on the GLOPS capability? I'm not very knowledgeable about video cards but a point of concern that I would have is the lower fillrate, memory bandwidth, and bus width of the GTX 650 compared to the GTX 560 Ti.

I have checked around and have not been able to confirm any specifications for the 600 series. If those specifications on Wikipedia are correct, I would say the GTX 650 would be an amazing card when it comes to performance given its power consumption. Any ideas on this?
2) Message boards : Number crunching : Prime number search (Message 38182)
Posted 3724 days ago by SiKTheGreatOne
I stand corrected. Thanks for clearing that up guys.
3) Message boards : Number crunching : Prime number search (Message 38145)
Posted 3725 days ago by SiKTheGreatOne
Prime numbers are only found from LLR applications. Factors are found from sieve applications. Factors are not prime and sieving helps by removing many candidates that are not prime. This helps because for an LLR task, the result is a prime or no prime, and sieving removes many of these no prime results. Anything that is not sieved must be LLR tested to determine if it is prime.
4) Message boards : Number crunching : My numbers don't add up! (Message 36512)
Posted 3767 days ago by SiKTheGreatOne
Free-DC updates at certain time intervals and will lag behind PrimeGrid, therefore your signature image and PrimeGrid stats may not match up exactly.
5) Message boards : Problems and Help : completed work not being submitted(win xp) (Message 28665)
Posted 3945 days ago by SiKTheGreatOne
Submission is automatic, it isn't instant (unless you create a cc_config.xml file for BOINC). Depending on your settings and the length of WUs you choose to work on, BOINC will automatically submit results at different times. BOINC will definitely submit your work before the WU deadline approaches however.

If you want instant reporting, you can read up on cc_config.xml here: http://boinc.berkeley.edu/wiki/Client_configuration.

The option you would want:

<report_results_immediately>0|1</report_results_immediately> If 1, each job will be reported to the project server as soon as it's finished, with an inbuild 60 second delay from completion of result upload. (normally it's deferred for up to a day, so that several jobs can be reported in one request). Using this option increases the load on project servers, and should generally be avoided. It's intended to be used on computer whose disks are reformatted every day.
6) Message boards : Problems and Help : completed work not being submitted(win xp) (Message 28660)
Posted 3945 days ago by SiKTheGreatOne
Open the BOINC manager, click the projects tab, and then select PrimeGrid. Click the update button on the left and you will find that your completed results will be reported.
7) Message boards : Number crunching : The Leonids Challenge (Message 28041)
Posted 3960 days ago by SiKTheGreatOne
Jim, if you only have 350 hours left between your two SOB tasks, you might want to try setting the work buffer to 10 days to see if you can get some SGS LLR work. (It should work since you said 350 hours total.) If the run time left for each task is 350 hours, then this will not work. Personally, I would recommend setting up a virtual machine using Ubuntu for the future. It's is very easy to do and comes in handy for challenges.
8) Message boards : Number crunching : Preparing for Leonids (Message 27553)
Posted 3974 days ago by SiKTheGreatOne
For about the last 24 hours, my hosts have been having intermittent trouble connecting to PrimeGrid. HTTP internal server errors, mostly. The problem is occurring right now and has been happening since about 12:15 UTC.

The stress test probably isn't going to mean much unless that problem is corrected first.


I've been getting plenty of HTTP internal server errors for over a week now. It happens on multiple PCs using two different internet connections.
9) Message boards : Number crunching : Sub-project "life" expectancy (Message 26867)
Posted 4001 days ago by SiKTheGreatOne
I assume the estimates are based on the average amount of work completed per day(perhaps per week or per month) in the past for a certain sub project as of recently and the life expectancy is calculated based on the LLR range/Sieve depth that is desired. Trying to accurately adjust the forecast for future hardware/rush of new users/hosts would be difficult and perhaps unnecessary. I believe that CPU manufacturers are aiming for more cache and more cores instead of higher clock speeds as more cores is an easier/better solution. GPU crunching on other sieves besides PPS is possible but the question of whether using a GPU for a particular sieve is worth it(does the GPU complete a WU in a satisfactory amount of time?). GPUs perform better on protein folding/MilkyWay than they do on PrimeGrid (GPUs are better suited for those other tasks). LLRing on a GPU may be possible, but again the question is: is the completion time of a task reasonable?
10) Message boards : Number crunching : X6 1055T - nice cruncher (Message 26436)
Posted 4021 days ago by SiKTheGreatOne
I put together a new build using the same chip recently as well. I have so far left mine at stock speeds and notice it is not that great with TRP Sieve(Still need to test with other subprojects). My Q9550 @ 2.83GHz completes TRP Sieve WUs in 3.5 hrs with little variation. The 1055T @ 2.8GHz can take anywhere from 4 hours to 5.5 hours. I'm not really sure why the completion times vary so much, but I may have to try OCing my chip as well.


Next 10 posts
[Return to PrimeGrid main page]
DNS Powered by DNSEXIT.COM
Copyright © 2005 - 2021 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 0.10, 0.04, 0.01
Generated 20 Sep 2021 | 23:20:28 UTC