The group's research activities comprise the
development and use of CFD-based analysis and optimization methods in
a variety of application domains. The majority of the routinely performed
computations are large scale ones and, thus, computationally demanding.
These are all performed on the high-performance computational platform
"VELOS" which comprises 3 clusters with 62 Teraflop computing power in total.
Each cluster consists of a number of interconnected CPUs (nodes)
employing the TCP/IP protocol. In particular:
The first cluster consists of 32 nodes (64bit, 120 cores, 110 GB in total).
Some nodes are equipped with NVIDIA GPU cards (GTX 280, 285, 580 & 670).
The second cluster comprises 70 DELL PowerEdge blade servers with
2 Quad, Six, Eight or Ten core Xeon GPUs each (64bit) 736 cores in total.
The RAM varies from 8GB to 128 GB per node, summing up to 3 TB for this cluster.
The third cluster is a GPU computing one which consists of 4 HP SL390s
servers and 4 Dell PowerEdge C8220X nodes with 12 Nvidia Tesla M2050
(3 on each HP server), 4 K20 and 4 K40 Nvidia GPUs with 104 GPU memory
in total.
The necessary services required for communication and data storage are handled
by two dedicated servers (one for the first two clusters and another for the
third one), with the Network Information Service (NIS, holding
the necessary user information) and Network File System Service
(NFS, for file sharing).
Time consuming numerical simulations employ the MPI protocol, according to the
MIMD (Multiple Instruction Multiple Data) model.
Parallel processing with the SIMT (Single Instruction Multiple Thread) model is
also performed using Graphics Processing Units (GPUs).
