Which BOINC project has lowest memory requerements?
17 Jan 2007 8:06:29 UTC
Topic 192321
(moderation:
Of the projects I take park in, my observations are that:
einstein allocates about 5Mb,
QMC allocates about 40Mb,
seti allocates about 50Mb,
abc allocates about 65Mb,
rosetta allocates about 100Mb
which means that einstein has the lowest memory requirements. Does someone know what the memory requirements are of the projects which I do not participate in?
"Everything should be made as simple as possible, but not simpler." A. Einstein
Which BOINC project has lowest memory requerements?
)
Malaria control showed up at around 10% on my laptop, so that would be around 50 MB, too. At the moment, though, my task manager says it's using 38 MB, maybe depending on WU size, or just because Windows and Linux show it differently.
One needs a bit more for CPDN, of course, I've seen that use up to 110 MB. Normally it's around 80 for me, though I've heard people complain about needing up to 200 even if they were only days into the model... might have been a faulty Windows installation or trouble with the BOINC client, though.
HashClash is down for the moment, but just for completeness' sake: around 110 MB there.
Primegrid takes about 16Mb
)
Primegrid takes about 16Mb here (with the primegen application), Chess960 should be at about 64Mb (I'm no quite sure about it, but it should be in that region) and SIMAP seems to take nearly 20Mb.
Here is table with memory
)
Here is table with memory requerements.
www.boinc.cz
Thanks, very nice table!
)
Thanks, very nice table!
"Everything should be made as simple as possible, but not simpler." A. Einstein
On my three G4 Macs the apps
)
On my three G4 Macs the apps that are currently in memory use approximately:
Einstein@home (v4.82): 6 MB
Leiden Classical (classical v5.36): 2.3 MB
SETI@home (optimized v5.13): 35 MB
SETI@home Beta (v5.17): 68 MB
SzTAKI Desktop Grid (v2.00): 1.0 MB
(As reported by Activity Manager under “Real memory�.)
Seti and SetiB will run in
)
Seti and SetiB will run in 32MByte, but will use more if it is available to speed up access to previously calculated variables. One of the reasons it works faster if there is lots of L2 cache. The scientific crunching part of program for ver 5.13 and 5.17 are the same, the main difference, besides some bug fixes, are to do with the new multi-beam receiver.