yeah i know what the projects are. i am being more specific:
Code:Making 1D domain decomposition 8 x 1 x 1
starting mdrun 'SINGLE VESICLE in water'
7250000 steps, 29000.0 ps (continuing from step 7087395, 28349.6 ps).
Printable View
yeah i know what the projects are. i am being more specific:
Code:Making 1D domain decomposition 8 x 1 x 1
starting mdrun 'SINGLE VESICLE in water'
7250000 steps, 29000.0 ps (continuing from step 7087395, 28349.6 ps).
It's a single viral particle in water.
Influenza virus is a vescile (membrane) encapsulated virus.
cool. i hope i dont have the flu. its going around my school like crazy.
I got my friend to try folding these big wu. For informational purposes his i7 at 3.6 ghz takes just under 34 minutes a percent under linux. He doesn't qualify for the bonus yet though.
i have noticed memory bandwidth makes a huge difference.
1070 cas6 it takes 45 minutes with my i7 3.76ghz
1400 cas8 it takes 36 minutes with i7 at 3.67ghz
No wonder, each of the 8 fah processes eats up over 400M right from the start. Whatever it's doing, it's memory intensive as hell. :eek:
what a beast of a simulation!Quote:
More information on these projects: they study the process of cell infection by the influenza virus (see description), which involves the fusion of liposomes (the fatty covering of the virus fusing with the fatty covering of the cell being infected). With over a million atoms in the simulation, one can easily understand the need for massive amounts of memory and computing power when processing bigadv units. The conventional SMP units are comparative featherweights, with a "mere" 147,087 atoms!
I only have 3Gb of memory in all but one machine it has 6gb. There only using 3.4Gb any way from what I can tell...
mine uses about 450 megs per thread. but my total ram usage is around 5 gigs which is the max i let it use.
Makes me think that these are purrfect for stability testing. 100% CPU with 2.10 core and loads of memory transaction tests for IMC stability. :)
Argh, I fold on my machines but they are all Windows. This doesn't work in any sort of VM right? I'd love to get the amount of ppd I'm getting total right now from a single cpu. I've got 2 systems I could run this on....argh.
It should run in VM if you can find one that supports 8 cores, I think there lies the problem... :shrug:
i think i turned it in after 4 days so no credit:shrug:. i didnt let it run for a day and for some stupid reason my other machine didnt have the passkey in it so i didnt have 10 wu's done. i have been trying to run it in windows through virtualbox but it crashes when it uses a lot of memory but it crashes a lot anyways. linux is pissing me off and i dont really want it to run on my main rig. i think im just going to buy some more gpu's and wait for smp2 or something.
well, I guess virtual box is the answer. Allows 64bit OSes with up to 32 processors. lets see about getting it to work without crashing...and still producing useful ppd.
Anyone have a guess on the upload and download sizes for these?
Out of interest, does anyone know how many cores bigadv supports? In case so want to use an 8-socket Istanbul machine for folding... :) :slobber:
I know 32 have been done for sure, and IIRC I think I saw results from a 64-core beast somewhere.
lol. when i run my gpu i get less ppd!:ROTF:my ETA in fahmon is only 7 hours from the deadline so i had to stop the gpu.
there are some 16 core opteron systems that are about as fast as road runners i7 @ 4.5 on foldingforum. you really have to go intel because gromacs utilizes sse. a 2p w3520 would dominate this thing for sure.
RR what happened?
its possible to get 10,000 PDD form a I7??? how??
will my PPD go up if i fold w/ 'large' size WU's on my Q9300?