The hypercube has been widely used as the interconnection network for parallel computers. However, in hypercubes, the number of communication links for each node is a logarithmic function of the total number of nodes. Therefore, the hypercube is not a good candidate for an interconnection network for a very large parallel computer that might contain hundreds of thousands of nodes due to IC technology and port number limitations. This paper introduces a new interconnection network for very large parallel computers called metacube (MC). An MC network has a 2-level cube structure. An MC(k, m) network can connect 2m2k+k nodes with m+k links per node, where k is the dimension of the high-level cubes (classes) and m is the dimension of the low-level cubes (clusters). An MC network is a symmetric network with short diameter, easy and efficient routing and broadcasting similar to that of the hypercube. However, an MC network can connect millions of nodes with up to 6 links per node. An MC(2,3) with 5 links per node has 16,384 nodes and an MC(3,3) with 6 links per node has 134,217,728 nodes. We describe the MC network's structure, topological properties and routing and broadcasting algorithms.
展开▼
机译：超立方体已被广泛用作并行计算机的互连网络。但是，在超立方体中，每个节点的通信链接数是节点总数的对数函数。因此，对于由于IC技术和端口号限制而可能包含成千上万个节点的大型并行计算机，超多维数据集不是互连网络的理想选择。本文介绍了一种用于超大型并行计算机的互连网络，称为 metacube （MC）。 MC网络具有2级多维数据集结构。 MC（ k，m ）网络可以将2 m2 k + k 个节点与 m +每个节点k 个链接，其中 k 是高级多维数据集（类）的维，而 m 是低级多维数据集（类）的维）。 MC网络是一种对称网络，具有直径短，类似于超立方体的简单高效路由和广播。但是，MC网络可以连接数百万个节点，每个节点最多可以有6条链路。每个节点具有5个链接的MC（2,3）具有16,384个节点，每个节点具有6个链接的MC（3,3）具有134,217,728个节点。我们描述了MC网络的结构，拓扑特性以及路由和广播算法。
展开▼
