DPAD: new version !!

[nukefarmer]

v4.4 is available from the DPAD homepage. This one also contains some new optimisations 🙂

get it while it's hot !

 

[Unforgiven]

here are the optimizations...

The major thing about this version is the release of two new optimisations: ChicaneLinacA, which is the continuation of the chicane design explored in versions 4.2x, and PhaseRotA, which includes the next stage of an alternative design intended to lead into a cooling ring later on.

The new straight sections include accleration via RF (radiofrequency) cavities, in which an electromagnetic field oscillates to accelerate the particles periodically. These are shown as orange/blue rings in the simulation. Since this introduces a time dependency into the tracking, the particles must be released from the initial rod with time-delays to model the timespread in the incident proton beam, which is why the particles now start off 'stuck' to the rod and only 'boil off' later.

The scoring system has had to be changed slightly for the new optimisations, as it is now such a difficult task that the initial random designs probably will have a transmission of zero, so instead negative scores based on how far the furthest muons have travelled have been assigned to the lower simulations, enabling them to climb towards achieving transmission.

Results are now 'compressed' into a binary format before being sent, to save some bandwidth (and also because the new optimisations have more parameters - up to 500 in some cases). This version also includes the fixes for memory leaks found in v4.34.

 

[Unforgiven]

does this mean we have to move over or can we still keep using 4.34? 😕

 

[GeoffS]

does this affect stats? If so, someone needs to contact me REALLY fast!

 

[Zbox]

yes indeed it affects stats 😀

a new v4.4 column has been added

 

[Zbox]

Originally posted by: Unforgiven
does this mean we have to move over or can we still keep using 4.34? 😕

you certainly don't have to move... v4.34 results are still being optimized

I've setup a few test systems to run v4.4 and I'll see how they go before rolling it out to the rest of my PCs. It seems that v4.4 will require a little more 'manual' configuring than the previous version as you will have to choose between the 3 sample files now available.

and on an interesting note, I've disappeared from Stephen's stats altogether 😛

 

[Unforgiven]

i use muoncalc to join results of all my machines to make manually uploading easier.... can you merge the 4.34 and 4.4 versions together or do they have to be kept seperate?

 

[Zbox]

I would keep them separate, as the new v4.4 results have many more variables and are compressed to a binary format before being submitted

 

[Unforgiven]

i guess ill just hold off until i really have to upgrade because hitting up the 6 boxes i have spread out here at work and home is a real pain. ill just sit tight until i hear the reviews on it 🙂

 

[Actaeon]

Thanks for the update nukefarmer!

 

[amdxborg]

lol This sux! Just this week I changed all my pcs at work from 4.33a to 4.34 after putting it off for some time... Well, looks like 30 something pcs needs to be changed again...🙁

 

[Zbox]

lol doh 🙂

 

[ProviaFan]

I've got a remote cow that I'm not sure if I'll be able to get to any time soon. 🙁

Hopefully when I do get to it I'll have time to get the new client installed...

 

[Zbox]

I would recommend that you stick with the previous client (v4.34) for those of you looking for maximum Mpts output. I've been running several tests and the older client is grossly faster at the SolenoidsTo15cm optimization. The performance of v4.4 actually decreases as the the simulation moves forward and I would be willing to say, at least right now, that v4.34 is anywhere from 3-5x faster (cumulatively).

If you want to delve into the new ChicaneLinacA & PhaseRotA optimizations, then install v4.4 by all means. Just realize that it will significantly alter your daily Mpts totals...

 

[Redviffer]

I'm not sure if I installed 4.4 correct, but now it will hang while attempting to download something. It's pissing me off because it doesn't ever download whatever it needs to. I manually shutdown and restart the client and it suddenly works again.

2 ways I've installed 4.4:

1) cleaned out the DPAD directory and installed all files in 4.4 into it.
2) extracted the new files in 4.4 into the existing DPAD directory, overwriting all files

Seems to be having the problem with both. Maybe I need to update the servers.txt ?


edit: yup, runs a lot slower also

 

[Zbox]

it's probably hanging on the lattice or sample file downloads

try setting both of these values in config.txt as follows:

Download sample results file after a number of days (0=don't): 0
Update lattice files from web every N hours (0=don't): 0

And you probably want to install it the first way you listed (by cleaning out the directory).

 

[Unforgiven]

the lattice issue is something that stephen plans to address in release 4.41 apparently (at least according to his post on his forum)

 

[Redviffer]

I went ahead and changed it like Z said:

Download sample results file after a number of days (0=don't): 0
Update lattice files from web every N hours (0=don't): 0


I'll change it back once it's been fixed. Thanks for the help. 🙂

 

[JonB]

In my Bug Report on the Muon forum, I reported that the "hang up" occurs every four hours. That comes from the default time for loading a new latticlist.txt file.

Just change the time to zero in config.txt and it works smoothly.

I saw that my client had produced no results overnight so looked in Win2k task managers.
Muon1 was still taking 99% cpu time, but NO RESULTS.

Then I noticed that the I/O Reads value was in the BILLIONS (to see I/O Reads, you have to select it in View/Select Columns. Somehow, it was never finishing the download segment for Latticelist.txt and just sat there. Since it actually did complete the download, when you restarted the client, it ran just fine for four hours, then it would hang again.

 

[Special1Sauce]

downloaded Dpadding

 

[Viztech]

I saw that my client had produced no results overnight so looked in Win2k task managers.

Interesting. I was having this trouble with the old 4.34 client on one machine only, so I upgraded it to 4.4 to check things out. It seemed to have downloaded everything fine on start-up and went to work.

The first chicane result had a negative value but the second a positive solenoids at 15 cm value. Does this sound normal?

viz

 

[Unforgiven]

if you havent upgraded to the new client yet, you may want to hold off. stephen (on his website forum) is speaking of optimizations to be released in the 4.41 client that will make the processing faster and comparable to version 4.34.

http://stephenbrooks.org/eve/ubb.x?a=tpc&s=724606111&f=3126012111&m=908109484

 

[Viztech]

Another fix due in ver. 4.41

Actually, latticelist.txt hung my Muon client too, so I'm going to fix this. It doesn't always do it though - on some networks it's fine. v4.41 should include a timeout.

I was having the same issue that JonB told about above, with both 4.34 and 4.4.

BTW, I just uploaded a 10.248146 result, crunched on a lowly P2-400! 😀

viz