“…The cloud can be modeled as a graph using the vertices as cloud server locations and the edges as the network latencies (Soyata, Friedman, & Mulligan, 1993) (Soyata, Friedman, & Mulligan, 1995) which will allow sophisticated combinatorial optimization algorithms for task scheduling in the cloud (Soyata, Friedman, & Mulligan, 1997) (Soyata T. , 2000). Additionally, rather than pure-software methods, hardware accelerators can be utilized for reducing the computational latency (Li, Ding, Hu, & Soyata, 2014) (Guo, Ipek, & Soyata, 2010) (Soyata & Liobe, 2012). While this will provide a speed-up and mitigate some of the negative impact of the network latencies, the next section focuses on the general effects of the latency on certain latency-sensitive applications.…”