Learn More
ÐRecursive Diagonal Torus (RDT), a class of interconnection network is proposed for massively parallel computers with up to P IT nodes. By making the best use of a recursively structured diagonal mesh (torus) connection, the RDT has a smaller diameter (e.g., it is 11 for P IT nodes) with a smaller number of links per node (i.e., 8 links per node) than those(More)
The interconnection network plays an important role in the performance and energy consumption of a network-on-chip (NoC) system. In this paper, we propose a RDT(2,2,1)//spl alpha/-based interconnection network for NoC designs. RDT(2,2,1)//spl alpha/ is constructed by recursively overlaying 2D diagonal meshes (torus). The number of layers needed for routing(More)
It has been well recognized that the fault-tolerance capability is vital for a NoC system, since one faulty link/processor may isolate a large fraction of processors. Continuing from a previous paper [13] where a RDT(2,2,1)/α-based interconnection network for NoC designs was proposed, in this paper, we investigate fault-tolerant routing schemes on NoC(More)
JUMP-1 is currently under development by seven Japanese universities to establish techniques of an ecient distributed shared memory on a massively parallel processor. It provides a memory coherency control scheme called the hierarchical bit-map directory to achieve cost eective and high performance management of the cache memory. Messages for maintaining(More)
In networks-on-chip (NoC) designs, delay variations and crosstalk noise have become a serious issue with the continuously shrinking geometry of semiconductor devices and the increasing switching speed. The crosstalk between adjacent lines causes data dependent signal delay and noise, thus finally makes the communication channel unreliable. The crosstalk(More)
The connection topology of interconnection networks and the routing algorithm often dominate system performance in MPCs. RDT has already been proved to be one of the best interconnection networks used in such systems. It supports a smaller diameter and degree than most at the size of one thousand to tens of thousands nodes. The floating vector routing(More)
In this paper, we present PRDT(2, 1), a new interconnection network topology for network-on-chip (NoC) design. PRDT(2,1) features a recursive structure, and has small diameter and average distance. We then focus our study on physical layout issues pertaining to PRDT(2, 1). Specifically, we show that the minimum number of metal layers required for the(More)
MapReduce is an effective programming model for large-scale data-intensive computing applications. Hadoop, an open-source implementation of MapReduce, has been widely used. The communication overhead from the big data sets' transmission affects the performance of Hadoop greatly. In consideration of data locality, Hadoop schedules tasks to the nodes near the(More)