> > >> A major advantage of these forthcoming ARM server systems is that >> they are air-coolable, and in fact even convection-aircoolable, >> if you add a suitable funnel on top of the rack. >> >> With SSDs, there are no movable parts. Float power will suck, >> but there are GPGPU options there, and memory bandwidth can >> be quite nice with memory cube like die stacking. > > > It would be lovely to have an article on clustermonkey for such a cluster. > At the very least it would make excellent teaching material [and for > poor-to-broken universities (and countries)] it would even make a putative > mini-research-cluster solution. Apologies for not offering to do it > myself, I wouldn't know where to start from ...
This may happen, however, a few sites you may want to look at first: 1. Little Fe (http://littlefe.net/home) 6 node dual core atom cluster. 2. Limulus Project (http://limulus.basement-supercomputing.com/) a 4 node with quad-core using Sand/Ivy Bridge processors (over 200 CPU-GFLOPS) in a single tower case. (disclosure: Limulus is my project). The software installs on top of Scientific Linux BTW. ARM may get real interesting as Project Denver (NVidia) develops. Of course, the AMD Trinity is a "similar" idea, i.e. SIMD/GPU co-processor on the CPU die. -- Doug (Cluster Monkey guy) > > > > > -- > > > Nicholas M. Glykos, Department of Molecular Biology > and Genetics, Democritus University of Thrace, University Campus, > Dragana, 68100 Alexandroupolis, Greece, Tel/Fax (office) +302551030620, > Ext.77620, Tel (lab) +302551030615, http://utopia.duth.gr/~glykos/ > > _______________________________________________ > Beowulf mailing list, [email protected] sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf > > -- > Mailscanner: Clean > -- Doug -- Mailscanner: Clean _______________________________________________ Beowulf mailing list, [email protected] sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
