Geert Dick van Albada

Computational Science – ICCS 2007

In the context of dynamic bandwidth allocation the QoS path provisioning for coexisted and aggregated traffic could be very important element of resource management. As we know all traffic in DiffServ/MPLS network is distributed among LSPs (Label Switching Path). Packets are classified in FEC (Forwarding Equivalence Class) and can be routed in relation to CoS (Class of Service). In the process of resource management we are looking for optimal LSP, taking care of concurrent flows traversing the network simultaneously. For better load-balancing purposes and congestion avoidance the LSP creation can be done off-line, possible during negotiation process. The main difference from well known routing techniques is that optimal LSP need not to be necessarily the shortest path solution as it is in the case of typical on-line routing (e.g. with OSPF protocol).

Computational Science – ICCS 2005

How can you change your mind to be more open? There many sources that can help you to improve your thoughts. It can be from the other experiences and also story from some people. Book is one of the trusted sources to get. You can find so many books that we share here in this website. And now, we show you one of the best, the computational science iccs 2005 5th international conference atlanta ga usa may 22 25 2005.

Computational Science - ICCS 2004

A 5-year-profiling in production mode at the University of Stuttgart has shown that more than 40% of the execution time of Message Passing Interface (MPI) routines is spent in the collective communication routines MPI Allreduce and MPI Reduce. Although MPI implementations are now available for about 10 years and all vendors are committed to this Message Passing Interface standard, the vendors’ and publicly available reduction algorithms could be accelerated with new algorithms by a factor between 3 (IBM, sum) and 100 (Cray T3E, maxloc) for long vectors. This paper presents five algorithms optimized for different choices of vector size and number of processes. The focus is on bandwidth dominated protocols for power-of-two and non-power-of-two number of processes, optimizing the load balance in communication and computation.