Filters








505 Hits in 4.9 sec

Computing the Optimal Longest Queue Length in Torus Networks [article]

Oscar Morales-Ponce and Burkhard Englert and Mehrdad Aliasgari
2021 arXiv   pre-print
We show that the algorithm attains the optimal queue length in O(σ C_max^2) rounds where C_max is the length of the longest conflict cycle with the maximum average queue length.  ...  Next, we present a centralized algorithm that computes a strategy in O(nlog n) time for each round that attains the optimal queue length in O(σ n) rounds where n is the number of nodes in the network and  ...  Minimizing the Longest Queue Length in Saturated Traffic Networks In this section, we present a strategy that minimizes the maximum queue length in saturated networks.  ... 
arXiv:1606.03800v2 fatcat:4y6p72rrkbhzfhqds23xbwv6nu

Application driven evaluation of network on chip architectures forcation parallel signal processing

C. Neeb, M. J. Thul, N. Wehn
2005 Advances in Radio Science  
We present a network-on-chip approach to derive an optimal communication architecture for a parallel Turbo-Decoder system.  ...  We focus on the strictly orthogonal n-dimensional mesh, torus and k-ary-n cube networks comparing deterministic dimension-order and partially adaptive negative-first and planar-adaptive routing algorithms  ...  For this, a (4,4)-2d-mesh network using input-queued routers is loaded with uniform traffic.  ... 
doi:10.5194/ars-2-181-2004 fatcat:tq5ustmoundzzaq666fu74njha

Improving parallel system performance by changing the arrangement of the network links

V. Puente, C. Izu, J. A. Gregorio, R. Beivide, J. M. Prellezo, F. Vallejo
2000 Proceedings of the 14th international conference on Supercomputing - ICS '00  
This network is an optimal 2D topology in the sense there are no other symmetric direct networks of degree 4 with a lower average distance or diameter.  ...  In fact, it reduces the diameter of the well known torus network by approximately √2 .  ...  network performance under uniform traffic with either a fixed or a bimodal (short and long messages) length distribution.  ... 
doi:10.1145/335231.335236 dblp:conf/ics/PuenteIGBPV00 fatcat:zpa7wqwgsbaedeitsd5djln5zy

Technology-Driven, Highly-Scalable Dragonfly Topology

John Kim, Wiliam J. Dally, Steve Scott, Dennis Abts
2008 2008 International Symposium on Computer Architecture  
By reducing global channels, a dragonfly reduces cost by 20% compared to a flattened butterfly and by 52% compared to a folded Clos network in configurations with ≥ 16K nodes.  ...  In this paper, we introduce the dragonfly topology which uses a group of high-radix routers as a virtual router to increase the effective radix of the network.  ...  This work has been supported in part by the National Science Foundation under Contract CCF0702341, in part by Cray, and in part by the Semiconductor Research Corporation under Contract SRC2007-HJ-1591.  ... 
doi:10.1109/isca.2008.19 dblp:conf/isca/KimDSA08 fatcat:oitsuv223jbodf255rctr5kezi

Technology-Driven, Highly-Scalable Dragonfly Topology

John Kim, Wiliam J. Dally, Steve Scott, Dennis Abts
2008 SIGARCH Computer Architecture News  
By reducing global channels, a dragonfly reduces cost by 20% compared to a flattened butterfly and by 52% compared to a folded Clos network in configurations with ≥ 16K nodes.  ...  In this paper, we introduce the dragonfly topology which uses a group of high-radix routers as a virtual router to increase the effective radix of the network.  ...  This work has been supported in part by the National Science Foundation under Contract CCF0702341, in part by Cray, and in part by the Semiconductor Research Corporation under Contract SRC2007-HJ-1591.  ... 
doi:10.1145/1394608.1382129 fatcat:lytq7l3yivdbrlzrfqypxf4jr4

Large-scale traffic signal control using machine learning: some traffic flow considerations [article]

Jorge A. Laval, Hao Zhou
2019 arXiv   pre-print
We are interested in the steady-state performance of the system, both spatially and temporally: we consider a homogeneous grid network inscribed on a torus, which makes the network boundary-free, and drivers  ...  We find that: (i) a policy trained with supervised learning with only two examples outperforms LQF, (ii) random search is able to generate near-optimal policies, (iii) the prevailing average network occupancy  ...  The signalized network corresponds to a homogeneous grid network of bidirectional streets, with one lane per direction of length n = 5 cells between neighboring traffic lights.  ... 
arXiv:1908.02673v1 fatcat:ngeffbvo2jcgbebfnbhi5zizbq

A novel optical network on chip design for future generation of multiprocessors system on chip

M. Channoufi, P. Lecoy, S. Le, R. Attia, B. Delacressonniere
2013 International Journal of Advanced Computer Science and Applications  
Keywords-3D-Optical network on chip, multi-level optical layer, optical control. I.  ...  Preliminary experimental results demonstrate the potential of multi-level optical layer for reducing power consumption and increasing scalability in the proposed ONoC.  ...  We first begin the performance evaluation study by demonstrating the potential of multi layer design in optimized standard network floorplans such as Mesh, Torus and passive lambda network.  ... 
doi:10.14569/ijacsa.2013.040336 fatcat:7hdfrbi6rbfndkklr76fgkscnu

Optimal Throughput–Delay Scaling in Wireless Networks—Part II: Constant-Size Packets

A. El Gamal, J. Mammen, B. Prabhakar, D. Shah
2006 IEEE Transactions on Information Theory  
While this tradeoff captures the essential network dynamics, packets need to scale down with the network size. In this "fluid model, " no buffers are required.  ...  In Part I of this paper, the optimal throughput-delay tradeoff for static wireless networks was shown to be , where and are the average packet delay and throughput in a network of nodes, respectively.  ...  Using the product form for the queue size distribution in equilibrium, it follows that the average queue size at a queue with total traffic and unit mean service is of the form where is some constant.  ... 
doi:10.1109/tit.2006.883548 fatcat:kmspyklnvzetjkhhsibip6mqmq

The BlackWidow High-Radix Clos Network

Steve Scott, Dennis Abts, John Kim, William J. Dally
2006 SIGARCH Computer Architecture News  
Each port consists of three 6.25Gb/s differential signals in each direction.  ...  By using a high-radix router with many narrow channels we are able to take advantage of the higher pin density and faster signaling rates available in modern ASIC technology.  ...  Acknowledgements Many people contributed to the development of the YARC router and BlackWidow network.  ... 
doi:10.1145/1150019.1136488 fatcat:746ci6xhgvd6bcqau4ifp45xo4

Small virtual channel routers on FPGAs through block RAM sharing

Jimmy Kwa, Tor M. Aamodt
2012 2012 International Conference on Field-Programmable Technology  
The throughput penalty from the neighbour traffic pattern can be reduced to 3% if a torus network is used. In all cases, there is little change in network latency at low load.  ...  This resource reduction comes at the cost of a 15% reduction in the saturation throughput for uniform random traffic and a 50% decrease in the worst case neighbour traffic pattern on a mesh network.  ...  In particular the neighbour traffic pattern can lead to two different extremes depending on whether the network is a mesh network or a torus network.  ... 
doi:10.1109/fpt.2012.6412115 dblp:conf/fpt/KwaA12 fatcat:l3o6c4oufbdvfeythgfza4g4xy

Deadlock-free XY-YX router for on-chip interconnection network

Yeong Seob Jeong, Seung Eun Lee
2013 IEICE Electronics Express  
In order to avoid deadlock, we add additional physical channels in the horizontal direction and optimize the priority of output channel allocation.  ...  Specifically, based on the observation that a response is always preceded by a request in multi-processor SoCs, this letter proposes a novel deadlock-free XY-YX router for on-chip network performance improvement  ...  Acknowledgments This work was supported in part by Seoul National University of Science and Technology, Korea and IDEC (EDA Tool).  ... 
doi:10.1587/elex.10.20130699 fatcat:jprfkps5b5fw5hhzrma3hsuspu

Centralized buffer router: A low latency, low power router for high radix NOCs

Syed Minhaj Hassan, Sudhakar Yalamanchili
2013 2013 Seventh IEEE/ACM International Symposium on Networks-on-Chip (NoCS)  
While router buffers have been used as performance multipliers, they are also major consumers of area and power in on-chip networks.  ...  At low loads, the CB is power gated, bypassed, and optimized to produce single cycle operation.  ...  ACKNOWLEDGEMENTS This research was supported in part by the National Science Foundation under grant CNS 0855110 and Sandia National Laboratories.  ... 
doi:10.1109/nocs.2013.6558397 dblp:conf/nocs/HassanY13 fatcat:agndczl4tjcr3kj2puaai5n2iu

Leaving One Slot Empty: Flit Bubble Flow Control for Torus Cache-Coherent NoCs

Sheng Ma, Zhiying Wang, Zonglin Liu Liu, Natalie Enright Jerger
2015 IEEE transactions on computers  
FBFC performs 92.8% and 34.2% better than LBS and CBS for synthetic traffic in a 4×4 torus. The gains increase in larger networks; they are 107.2% and 40.1% in an 8×8 torus.  ...  Short and long packets co-exist in cache-coherent NoCs. Existing designs for torus networks do not efficiently handle variable-size packets.  ...  For synthetic traffic, FBFC performs 92.8% and 34.2% better than LBS and CBS in a 4×4 torus. FBFC's advantage is more significant in larger networks; these gains are 107.2% and 40.1% in an 8×8 torus.  ... 
doi:10.1109/tc.2013.2295523 fatcat:gtlcmsgftjcjtnqrz2btntmgdu

SB-Router: A Swapped Buffer Activated Low Latency Network-on-Chip Router

Monika Katta, T K Ramesh, Juha Plosila
2021 IEEE Access  
ONLINE SB Swapping the contents of the entire VC is not an optimized design. The associated circuitry and its complexity increase with the increase in the length of SB.  ...  With the PARSEC benchmark, the average network latency of the baseline router is compared with SB-Router for the mesh network and the results are shown in Fig. 18 . E.  ...  the cyber-physical systems and internet-of-things domains, with a recent focus on fog/edge computing (edge intelligence), and autonomous multidrone systems.  ... 
doi:10.1109/access.2021.3111294 fatcat:5wnkftpzpfahxgyrcpzxvj4bri

Table-lookup based Crossbar Arbitration for Minimal-Routed, 2D Mesh and Torus Networks

DaeHo Seo, Mithuna Thottethodi
2007 2007 IEEE International Parallel and Distributed Processing Symposium  
Crossbar arbitration-which determines the allocation of output ports to packets in the input queues-is a performance-critical stage in the overall performance of routers for input-queued networks.  ...  networks.  ...  This work is supported in part by NSF grant CCF-0541385, Purdue Research Foundation XR Grant No. 6904010 and Purdue University.  ... 
doi:10.1109/ipdps.2007.370272 dblp:conf/ipps/SeoT07 fatcat:xy4jxybavjbphpzv4hlqz7dcqm
« Previous Showing results 1 — 15 out of 505 results