Communication latency has become one of the determining factors for the performance of
parallel clusters. To design low-latency network topologies for high-performance computing …

This article considers the usage of circulant topologies as a promising deadlock-free
topology for networks-on-chip (NoCs). A new high-level model, Newxim, for the exploration …

Industry 4.0, smart cities, Internet of Things, big data, and digital transformation are the main
paradigms and technologies of the artificial intelligence era. This era requires a foundation …

摘要以机器学习为代表的人工智能技术需要对海量数据进行处理, 对底层算力要求极高.
分布式机器学习通过将计算任务分布式地部署到多个计算节点来加快模型的训练速度 …

This article presents software for the synthesis of circulant graphs and the dataset obtained.
An algorithm and new methods, which increase the speed of finding optimal circulant …

This article considers the usage of circulant topologies as a promising deadlock-free
topology for networks-on-chip (NoCs). A new high-level model, Newxim, for the exploration …

A network is connected if there exists a path between any two distinct vertices. The vertex-
connectivity of any connected network is the cardinality of its minimum vertex-cut. Then, a …

Using three supercomputers, we broke a record set in 2011, in the enumeration of non-
isomorphic regular graphs by expanding the sequence of A006820 in the Online …

Parallel versions of a genetic algorithm based on the hybrid MPI—OpenMP model are
implemented to optimize circulant networks, which are of practical interest in the design of …

Training artificial neural networks can be discrete (digital) and continuous (analogue). In the
former case, machine learning has always been the sphere of search for ways to optimise …