C. E. Leiserson. Fat-trees: Universal networks for hardware-ecient supercomputing. IEEE Transactions on Computers, C-34(10):892-901, Oct. 1985.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....analysis. We justify the accuracy of the approximation by means of behavioral simulations based on a gate level implementation of a fat tree network. 1 Introduction Fat tree networks are established as area universal communication networks due to the seminal work of Charles E. Leiserson [8, 3], culminating in the implementation of the Connection Machine CM 5 at Thinking Machines Corporation [9] Today, advances in semiconductor technology enable us to integrate multiprocessor machines on a single chip, as explored in the Raw project [12] for example. As the number of processors on a ....

....randomly chosen sources and destinations can be delivered in a fat tree network with n processing nodes within O(lg m) delivery rounds with high probability. Our result improves on previously published bounds based on the number of processing nodes n rather than the number of messages m. Leiserson [8] derived a bound using the load factor of a set of messages. He has shown that the number of delivery rounds required to deliver a set of messages, where the sources and destinations are known in advance, is O( lg n) Greenberg and Leiserson [3] have derived a bound O( lg n lg lg n) for the ....

Charles E. Leiserson. Fat-Trees: Universal Networks For Hardware-Ecient Supercomputing. IEEE Transactions on Computers, C-34(10):892-901, October 1985.

....analysis. Some concluding remarks are given in section 6. 2 The QsNET The QsNET is based on two building blocks, a programmable network interface called Elan [12] and a lowlatency high bandwidth communication switch called Elite [13] Elites can be interconnected in a fat tree topology [7]. The network has several layers of communication libraries which provide trade o s between performance and ease of use. Other important features are hardware support for collective communication patterns and fault tolerance. 2.1 Elan SDRAM I F Processor code DMA Buffers Inputter FIFO ....

Charles E. Leiserson. Fat-Trees: Universal Networks for Hardware Ecient Supercomputing. IEEE Transactions on Computers, C-34(10):892-901, October 1985.

....as the number of nodes increases, as do unidirectional multistage networks. 1 Note: Switch2 is not an IBM product and no assumptions should be made regarding its availability in the future. ADAPTIVE ROUTING ON SP SWITCH 3 They have smaller diameter than the mesh and torus networks. Fat trees [18] and least common ancestor networks [21] are examples of BMINs. In the source routing mechanisms the network interface must perform a route table lookup before inserting messages into the network. Route tables are built by various algorithms that use the network topology as an input. In the ....

....chip in the network. As do uni directional mul 6 ABALI et al. tistage networks, these networks scale aggregate bandwidth linearly as the number of nodes increases. BMINs reward communication locality, and require no virtual channels to avoid deadlock among packets in the network. Fat trees [18] and least common ancestor networks [21] are examples of BMINs. Figure 2 shows the path of two packets sent from node P13. The rst message is destined for P8, and need only be routed through one switch chip. The second message is destined for P6, and traverses 3 switch chips. Note that any of ....

C. E. Leiserson. Fat-trees: Universal networks for hardware-ecient supercomputing. IEEE Trans. on Computing, C-34(10):892-901, Oct. 1985.

No context found.

C. E. Leiserson. Fat-trees: Universal networks for hardware-ecient supercomputing. IEEE Transactions on Computers, C-34(10):892-901, Oct. 1985.

No context found.

C. E. Leiserson. Fat-trees: Universal networks for hardware-ecient supercomputing. IEEE Transactions on Computers, C-34(10):892-901, Oct. 1985.

No context found.

C. Leiserson, "Fat-trees: Universal Networks for Hardware-e#cient Supercomputing," IEEE Transactions on Computers, vol. 34, no. 10, pp. 892-901, October 1985.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC