New lattice for DPAD PhaseRotEby5

[waffleironhead]

Stephen has released a new lattice "PhaseRotEby5" which is replacing PhaseRotDD. While the client is supposed to handle the new lattice changeover, its usually best to make sure everything is going ok.

 

[petrusbroder]

Thanks for the heads-up, waffleironhead! :sun:

 

[Assimilator1]

Thanks waffle

Will remember that when I fire backup 2 DPAD clients next week.

What is a lattice anyway?

 

[waffleironhead]

I dont know what is really contained in the lattice exactly. But it is used to interact with the client and tell it exactly what simulation to run. I think it is a set of instructions or parameters.

this is what is contained in a lattice file.

as im not a programer, its all foreign to me

 

[stephenbrooks]

You'll be lucky if you can understand that file by eye. It's a programming (scripting) language I wrote specifically for Muon1 and designing particle accelerators. It also has a pseudo-C interpreter bashed together with it.

I posted an explanation of the lattice and its progress in the Muon1 forum here.

Just today I played with the stats tables slightly. Mostly, it means the all team stats page is improved a bit. I noticed Team Anandtech wasn't far off DPC, wondered if you were planning on overtaking!

 

[Orange Kid]

Just today I played with the stats tables slightly. Mostly, it means the all team stats page is improved a bit. I noticed Team Anandtech wasn't far off DPC, wondered if you were planning on overtaking!

Trying to stir the pot a bit

As to the lattice file , mine keep trying, but fail. I'm guessing that traffic is heavy and it'll get it sooner or later

 

[caferace]

Originally posted by: stephenbrooks
You'll be lucky if you can understand that file by eye. It's a programming (scripting) language I wrote specifically for Muon1 and designing particle accelerators. It also has a pseudo-C interpreter bashed together with it.

I posted an explanation of the lattice and its progress in the Muon1 forum here.

Just today I played with the stats tables slightly. Mostly, it means the all team stats page is improved a bit. I noticed Team Anandtech wasn't far off DPC, wondered if you were planning on overtaking!

whoa. Hi Stephen. I didn't know you dropped by here. ::waves::

-jim

 

[Orange Kid]

Ok, so still not getting the file, so I dl'd it manually and put it in the samples folder.
Seem to be running fine, we'll see

 

[stephenbrooks]

Hi caferace, yes, I did used to check some of these team forums more often but haven't done in what seems like ages. I've found that people are sometimes happier discussing stuff on these boards than in my forum as well.

Orange Kid -- I'm not sure what the file problems are. The file you want is PhaseRotEby5.txt and it should go in the lattices directory NOT the samplefiles one. There are no samplefiles available for this lattice at this early stage, so don't worry if it tries and fails with that.

If it's contacted the server OK, then you should find PhaseRotDD.txt has disappeared from your lattices folder too.

As for "stirring it", I'm just saying, DPC seems to have been sitting at the top for rather a long time

 

[Assimilator1]

stephenbrooks
Hi ,good to see ya here

We were catching DPC a little while back ,not sure what the current state of play is,but hopefully we'll catch 'em soon

 

[Orange Kid]

The PhaseRotEby5.txt is in the lattices directory and running fine. I also put an empty PhaseRotEby5.txt in the sample directory to stop the client from complaining on a machine was giving me grief and this seemed to satisfy it.

I myself like to 'stir the pot' and really wish more people here would do this project. Not just to give the DPC some competition, but because I feel it is one of the more useful and relevant projects around. I had stopped for a while because the 'c' client was crashing a lot on my pc's but the 'd' client is nicely behaved for me.

 

[stephenbrooks]

I just got to work and found the problem - it wasn't refreshing the samplefiles with the right filenames (I convert most filenames that are going onto the web to lowercase) so it was looking for "phaseroteby5" when the file was actually called "PhaseRotEby5". This problem doesn't apply to the lattices, only the samplefiles, and now appears to be fixed.

What's interesting is how up until now that means every person (well, apart from those who manually use the sample) has gone on a separate optimisation path, in some cases all the way up to 0.5%.

 

[Assimilator1]

I hear ya Orange Kid ,would be nice to crank up TA DPAD ,I think part of the trouble is that people have trouble relating to what the project is about & also what's different about it compared to LHC (can't remember that one myself now!).

It's a pitty that the part about hoping to transmute radiocative waste isn't more ....err active!.I know research was picking up on that again after some years of apparent silence but I have no idea what it is atm.Anyway ,people could easily see the uses of that part of it.

 

[stephenbrooks]

The LHC project was quite odd as they'd already built a load of magnets that they were going to use in the tunnel but wanted to know what was the best order to put them in (having tested them all and recorded each one's particular errors).

But in the scheme of things that's a relatively minor variation of design, Muon1 of course can do much bigger variations as the machine itself is in the design stage rather than about to be switched on.

Just recently there's been another resurgence of interest in building the muon collider in the US, which would have an initial stage in common with the neutrino factory and the simulations of the part of it I'm doing. The muon collider could be a successor to the LHC, it would have a very diverse physics reach.

By the way, does waffleironhead know how he posted the very first result in PhaseRotEb6 and managed to go positive immediately? That wasn't a seeding of it with the other optimisation was it?

 

[waffleironhead]

yeah it was a run from the eby5, I just changed the lattice it was on and reran it. sorry if it causes problems.

 

[stephenbrooks]

OK, not a problem really, it seems to have started improving upon your result already!

 

[Assimilator1]

Interesting info (funny that LHC only ran DC near the end of the build ),though I wasn't only refering to the clients goals but also the final hardware's goals.I only vaguely recall that different research will be carried out but can't remember what

 

[stephenbrooks]

The LHC is a collider of protons with protons, at the high energy (will be the highest ever achieved) of 7TeV per beam. This is a search for exotic particles at energy scales currently inaccessible at lower energies.

The neutrino factory has a different problem: neutrinos don't require a high energy to be produced, but they have what is called a low "interaction cross-section", meaning that they can pass through an enormous quantity of material (e.g. the whole Earth) with only a small likelihood of actually colliding with anything. Thus they are difficult to manipulate in conventional ways. We need an accelerator to accelerate the source of neutrinos (muons that decay) in a specific direction so that the neutrinos inherit their momentum and form a beam rather than spraying everywhere. We can then put detectors at other places on the Earth to intercept the beam when it resurfaces.

(7TeV is an energy of 1.12 microJoules per individual proton. That's the energy of a 1 gram marble travelling at nearly 5cm/s, i.e. enough to feel, in a single subatomic particle. And the LHC has billions of them going round.)

 

[Assimilator1]

Thanks Stephen

And what will we be hoping to find out (directly & indirectly) from detecting the Neutrino beam?

(I'm sure I used to know the answer to that one once! lol)

 

[stephenbrooks]

The main thing they want is a more precise determination of the three neutrino masses. We have a better idea of the masses of the other 9 particles of matter, but neutrinos weren't even known to have any mass until the 1990s.

The interesting thing is how they have very small masses, at a different scale from anything else. That big gap in scale is interpreted as an important clue in trying to come up with unified physics theories, where the masses are predicted from something more fundamental.

There are also some slightly more complicated quantities to do with how the neutrinos interact with different sorts of matter particle. Those are also of interest for theorists, because they can be compared to the quark family, also consisting of 3 generations of particle (2 quarks in each).

Exactly how likely it is they'll find a theory that fits, I don't know, but finding these quantities with precision rules out a lot of theories, hopefully shedding some light on the remainder.

 

[Assimilator1]