• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

DPAD: need some details please..

C

CyGoR

Platinum Member
When starting the DPAD client it needs to download a new sampleresults file (excisting one is older then 72hrs)
but it prints out this line 20 times or so: Record rejected due to bad checksum: was {77F421E0}...
What does this mean for the progress? The client runs fine by the looks of it..

Also, I've got a PIV 2.0Ghz which isn't on the network and does not have a floppy drive (I swap HD's sometimes) which
was running Seti (huge buffer😉) but I want it to run DPAD, any ideas on how to set it up? I can download the client, but how can
I make sure it has enough work for let's say a week?

 
Are you using v4.33a or 4.34? There's no real problem with that, just that the 4.33a client doesn't check the checksums correctly, IIRC 🙂



And for the other computer, DPAD doesn't have "work units". It can crunch offline for as long as you want 🙂 Set the download lattice list and download sample results to = 0 and it will be fine 🙂 When you want to dump, just copy the results.txt file to another PC which has internet access and run manualsend.bat 🙂


Garry
 
not to hijack this thread but what is the purpose of the lattice and sample files? ive always had mine set to 0 since i started 2 weeks ago. could someone clue me in on what the purpose of those files are and is it benefical to have it set to 0, yes or no? also, confused it right bro, just fire up dpad and let it crunch for a whole year if you wanna without uploading! thats one of the things that is great about this project and the top reason im doing it!
 
Thanks Garry! 🙂

I don't know which version I use, the readmy file says 4.3X 😉 I though that I'm using 4.33a, downloaded the executable installer from the FAQ.

Youre not hijacking this thread Unforgiven, those same questions were crossing my mind as well.

Here's a piece of my config.txt file, the obvious things (resolution) aren't pasted..
------------------------------------------------------------------------------------------------------
Auto-send results ([Y]es, [N]o): Y
Auto-save interval (seconds; 0 for no save): 230
Rechecks for best-so-far results (min. 5): 5
Use passive (PASV) mode for FTP transfers: Y
Download sample results file after a number of days (0=don't): 3
Sample file URL: http://stephenbrooks.org/muon1/sample100.txt
FTP command call: ftp -s:script.ftp

I've got the feeling I'm missing something here 😉
 
Yep, you're V4.33a 🙂

The latticelist is a list of the files in the lattices folder. It makes sure that you've got the right list 🙂

The sampleresults file is the top 100k or 500k worth of results. The program works by trying to improve the best design. Of course, if it's got a list of the best designs so far, then it has a good basis to go from! The advantage of using it is that you will get higher transfer value, and are more likely to get more Mpts per simulation.

If you don't download the sampleresults, then the program will start from scratch, and will take different approaches. This could yield better, or worse, results than using the sampleresults file 🙂 Either way, it is user preference 🙂


Garry
 
ok well this spawns another question for me. say that i set both of those settings to 0 so essentially im starting from scratch with my setup....does my dpad installation teach itself what methods its already done and over time yield me higher points/results based on what im doing locally on my pc or will the numbers remain low and never get higher without me downloading the sample results file? ive noticed lately an increase (based on the numbers im seeing from using muon calc) in the number of points im getting overnight as opposed to when i was first getting started. im not sure whether this is coincidence or if there if its learning a better way to crunch so its starting to get higher results...any thoughts would be greatly appreciated!
 
The program will learn and improve, be it from what it's done itself in the past, or from the sample file downloaded from the net 🙂


Garry
 
Thanks again 🙂

I'll copy the files and let it run for a week or so.. Friday is flush day? 😉
 
awesome garry, thank you so much for the information! i dunno about friday being dump day but i just unloaded a HUUUUUUGE quantity from overnight crunching! 🙂
 
Just for u guys to know: The url in the line "Sample file URL: http://stephenbrooks.org/muon1/sample100.txt" isn't working anymore!! Ig u use 4.33a and want to download the samplefiles change it to "http://www.stephenbrooks.org/muon1/samplefiles/solenoidsto15cm_100.txt". Stephen has renamed the file! 😀

Btw, I just set up a laptop with 4.34 and I also downloaded the sample files but still this laptop is giving me 500 results with no transfer higher than 0.6 and no more than 30 mpts per run! Any ideas on that?
 
Sorry BMC, can't help you on that one, I don't have any knowledge about output from DPAD to say what is normal and what isn't..
But thanks for your help! Changed the file name in the config file 🙂
 
i dont download the sample files but my laptop yields horrible results. i get an occasional one over 100 pts but other than that it averages 25-50 pts per unit. its a 433 celeron so its not the fastest cpu (obviously) but a PII 450 in the same office nets huge point totals over 500 at some points. how much is the points system effected by cpu speed? i mean is the software learning on its own or are you bottlenecked by the cpu?
 
CPU only affects the speed it crunches at...Redownload the best results file and see what happens. Make sure to kill the current process and restart it.
 
I downloaded the whole DPAD package 3 times so far and it's the same thing everytime. I really don't know what's causing the problem. The interesting thing is the following:

I took my main crunchers DPAD folder and copied it to my laptop. With all samples and .dat files. This DPAD folder was constantly giving me 300+ runs. But as soon as on the laptop, it only gives me 30 mpts runs. It's a PIV 2.8 HT laptop. I really have NO idea what this is about. I think it's kinda funny actually. 😀
 
yeah im not too hip on the scoring details but i have a few P4 HT boxes here that get worse results than my laptop but then kick out more units so the total points are really higher in the end due to the cpu speed 🙂 i dunno whats going on either but i know im staying the course and as long as im getting production i dont care how small the points are 🙂

i just appended the 500 sample results file to the end of my results.dat and im gonna see if this makes any difference at all in my points collection 🙂
 
I guess we need to know what CPU make, model, stepping #, etc... you are having results problems with. The genetic algorithm that DPAD uses to adjust itself may be having problems with certain CPU architectures. DPAD probably uses random number generation and seed values and different processors can give strange results.

My A64 3000+ gives me "checksum" errors when it tries to download a new sampleresults file, but gives me very consistent sets of results near 300 MPts.

My laptop's Athlon XP 1700 doesn't seem to have checksum errors, though, and it also gives near 300 MPts results.

both are running 4.34v

 
after appending the 500 results sample file to the end of my results.dat file my laptop has been doing units consistently over 300 points 🙂 this made a HUGE difference and ive applied the same thing to all of my crunchers for this project to see if the results are the same. bmc you might wanna try the same thing if you dont have your sample results to automatically update.

heres how to do it (right of stephen brook's faq) if anyone is interested and doenst have their config.txt file setup to automatically check for them:

How do I use the sample file?
Decompress it if necessary and either use it to replace your results.dat, or edit your results.dat in WordPad and append the new results (being careful not to merge lines). The latter option will mean Muon1 can use your old results as well as the downloaded ones, so is a better choice.

**note**
make sure you follow the directions exactly and use wordpad and NOT notepad. notepad created line breaks for me and threw my results.dat all off, so make sure to use wordpad.
 
Well, the problem somehow solved itself cause I've had to give away that laptop again 🙁 When I get my hands on it again I'll try what u suggested Unforgiven! 😀
 
yeah man, all my machines are pumping out 300+ point units now 🙂 thats just a cross section from the 4 machines im running here at work for the past 2 hours!
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top