2,429 Hits in 1.8 sec

Flit-reservation flow control

Li-Shiuan Peh, W.J. Dally
Proceedings Sixth International Symposium on High-Performance Computer Architecture. HPCA-6 (Cat. No.PR00550)  
This paper presents flit-reservation flow control, in which control flits traverse the network in advance of data flits, reserving buffers and channel bandwidth.  ...  Flit-reservation flow control requires control flits to precede data flits, which can be realized through fast on-chip control wires or the pipelining of control flits one or more cycles ahead of data  ...  Current flow-control methods are reviewed and compared to flit-reservation flow control in Section 2.  ... 
doi:10.1109/hpca.2000.824340 dblp:conf/hpca/PehD00 fatcat:sejwtfrm6jemrmabdxkixocx7y

Runtime connection-oriented guaranteed-bandwidth network-on-chip with extra multicast communication service

Faizal Arya Samman
2014 Microprocessors and microsystems  
The routing is organized using runtime programmable routing reservation table. In addition, the proposed methodology supports also a deadlock-free multicast routing service.  ...  The contention problem is solved with the hardware solution based on the locally organized message identity (ID), in which flits belonging to the same stream packet will have the same unique/local identity-tag  ...  By using this technique, a message is not split into packets, but it is split into flow control digits (flits) with extra bit fields for dynamic unique/local ID and flit-type control label per flit.  ... 
doi:10.1016/j.micpro.2013.07.006 fatcat:myvqukymofbezfmqmtciybnfpi

Switching Techniques [chapter]

Jack Dongarra, Piotr Luszczek, Felix Wolf, Jesper Larsson Träff, Patrice Quinton, Hermann Hellwagner, Martin Fränzle, Christian Lengauer, Luis H Ceze, Kei Hiraki, Rolf Riesen, Arthur B. Maccabe (+47 others)
2011 Encyclopedia of Parallel Computing  
SYNONYMS Flow Control DEFINITION Switching techniques determine how messages are forwarded through the network.  ...  Flit reservation flow control was introduced to improve buffer turn around time in wormhole switching [16] .  ...  Many schemes have been developed over the years for flow control including on-off, credit-based, sliding window, and flit reservation.  ... 
doi:10.1007/978-0-387-09766-4_296 fatcat:eudzsnjzzvgwdjft2pikws4hkq

Wormhole cut-through switching: Flit-level messages interleaving for virtual-channelless network-on-chip

Faizal Arya Samman, Thomas Hollstein, Manfred Glesner
2011 Microprocessors and microsystems  
For quality of service purpose, inter-switch data transfers are controlled by using link-level overflow control to avoid drops of data.  ...  Each flit belonging to the same message can track its routing paths correctly because a local identity-tag (ID-tag) is attached on each flit that varies over communication resources to support the wire-sharing  ...  Link level overflow control and flit-by-flit arbitration Data flows in our NoC are controlled using a link-level flit flow control.  ... 
doi:10.1016/j.micpro.2011.01.003 fatcat:j6nhztnl2vcj5e2vgofsijamja

Dynamic Channel Flow Control of Networks-on-Chip Systems for High Buffer Efficiency

Sung-Tze Wu, Chih-Hao Chao, I-Chyn Wey, An-Yeu Wu
2007 Signal Processing Systems Design and Implementation (siPS), IEEE Workshop on  
Moreover, the throughput can be improved by 20% as compared with wormhole flow control.  ...  By the experimental result, the proposed mechanism can reduce the buffer size by 30% as compared with virtual channel flow control at the same throughput.  ...  Control flits traverse the network before the data flits to reserve buffers of the channel. Data flits are forwarded or buffered through the reserving channels.  ... 
doi:10.1109/sips.2007.4387597 dblp:conf/sips/WuCWW07 fatcat:lwsiy4us5fh5legrl6c6bj7yhq

A conflict sense routing protocol and its performance for hypercubes

E.A. Varvarigos, D.P. Bertsekas
1996 IEEE transactions on computers  
Following the terminology of [Dal90] we define a flit (from "flow control unit") as the smallest number of bits which can contain routing information, or else the minimum number of bits which can be accepted  ...  We assume that a node has a way to distinguish control flits from actual data.  ...  flow control mechanism becomes easy and efficient.Note that the control flits can also be transmitted "off" channel.  ... 
doi:10.1109/12.506425 fatcat:jbtmrfdpwfbrvm7d2x7ifkcvra

Network-on-Chip with load balancing based on interleave of flits technique [article]

Marcelo Daniel Berejuck
2015 arXiv   pre-print
The NoC is based on a technique that allows the interleaving of flits from diferente flows in the same communication channel, and keep the load balancing without a centralized control in the network.  ...  Analytic comparisons and experimental data show that the approach adopted in the network keeps average latency lower for variable bitrate flows than a network based on resource reservation when both networks  ...  Second, it offers best-effort services allowing the flows treated as BE to use reserved, but unused, time slots time slots to improve the throughput of these flows.  ... 
arXiv:1510.06791v1 fatcat:5g3edc2ksvdhhpxor5xiwrnscy

The efficiency of buffer and buffer-less data-flow control schemes for congestion avoidance in Networks on Chip

Ahmed Aldammas, Adel Soudani, Abdullah Al-Dhelaan
2016 Journal of King Saud University: Computer and Information Sciences  
It describes the internal functions of this router for optimal output flit scheduling and its capability to apply per-class service for inbound flows.  ...  This paper proposes an internal router architecture, for on chip communication, implementing flow-control mechanism for congestion avoidance with QoS consideration.  ...  The authors in Peh and Dally (2000) suggest a data-flow control process using a flit-reservation mechanism that reserves buffer space and link bandwidth for the data stream.  ... 
doi:10.1016/j.jksuci.2015.11.002 fatcat:kqtl6lmhijakxlrp7eylz6cn4e

Evaluation of silicon consumption for a connectionless Network-on-Chip [article]

Marcelo Daniel Berejuck, Antônio Augusto Fröhlich
2014 arXiv   pre-print
The design is based on a connectionless strategy in which flits from different communication flows are interleaved in the same communication channel between routers.  ...  Analytic comparisons show that our approach keeps average latency lower than a network based on resource reservation, when both networks are working over 80% of offered load.  ...  The flow controller checks the destination fields from the flit stored in the input interface register and informs to another block called routing controller about the destination requested in the flit  ... 
arXiv:1411.3492v1 fatcat:csm2pcjb2zb4dkd2ibztja5woq

Network-Cognitive Traffic Control: A Fluidity-Aware On-Chip Communication

Wen-Chung Tsai, Sao-Jie Chen, Yu-Hen Hu, Mao-Lun Chiang
2020 Electronics  
A novel network-on-chip (NoC) integrated congestion control and flow control scheme, called Network-Cognitive Traffic Control (NCogn.TC), is proposed.  ...  An integrated congestion control and flow control algorithm is proposed based on the estimated buffer fluidity level.  ...  It realizes flow control on a new group that comprises several flits.  ... 
doi:10.3390/electronics9101667 fatcat:ry4yavratja5lhypnezvnpyayu

Channel reservation protocol for over-subscribed channels and destinations

George Michelogiannakis, Nan Jiang, Daniel Becker, William J. Dally
2013 Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis on - SC '13  
Furthermore, ECN may not prevent congestion formed by short-lived flows generated by a large combination of source-destination pairs.  ...  We propose channel reservation protocol (CRP) to enable sources to reserve bandwidth in multiple resources in advance of packet transmission and with a single request, but without idling resources like  ...  Packet size is set to 256B, divided into 32 8B flits. Control packets consist of one 8B flit. Sources generate traffic in units of messages (flows).  ... 
doi:10.1145/2503210.2503213 dblp:conf/sc/MichelogiannakisJBD13 fatcat:k6c67eq5o5ghbiwki64rpfsdsi

MMR: a high-performance MultiMedia Router-architecture and design trade-offs

J. Duato, S. Yalamanchili, M.B. Caminero, D. Love, F.J. Quiles
1999 Proceedings Fifth International Symposium on High-Performance Computer Architecture  
The distinguishing features of the proposed router architecture are the use of small fixed-size buffers, a large number of virtual channels, linklevel virtual channel flow control, support for dynamic  ...  The MMR uses flow control to prevent flits from being discarded. Additionally, flit buffers are relatively small, producing a fast propagation of flow control information.  ...  The MMR uses flow control to prevent flits from being discarded. Additionally, flit buffers are relatively small, producing a fast propagation of flow control information.  ... 
doi:10.1109/hpca.1999.744383 dblp:conf/hpca/DuatoYCLQ99 fatcat:7eczzvyo5jcztkjld6lh7ajm4y

Worm-Bubble Flow Control

Lizhong Chen, T. M. Pinkston
2013 2013 IEEE 19th International Symposium on High Performance Computer Architecture (HPCA)  
In this paper, we propose Worm-Bubble Flow Control (WBFC), a new flow control scheme that can avoid deadlock in wormholeswitched tori using minimally 1-flit-sized buffers per VC and one VC in total.  ...  Deadlock-free flow control should be designed with minimal cost, particularly for on-chip designs where area and power resources are greatly constrained.  ...  In this paper, we propose Worm-Bubble Flow Control (WBFC), a new flow control scheme that can avoid deadlock in wormhole-switched tori using minimally 1-flit-sized buffers per VC and one VC in total.  ... 
doi:10.1109/hpca.2013.6522333 dblp:conf/hpca/ChenP13 fatcat:zotvjad3kfhyvpkl7ulcxd7dze

Advanced Design Issues for OASIS Network-on-Chip Architecture

Kenichi Mori, Adam Esch, Abderazek Ben Abdallah, Kenichi Kuroda
2010 2010 International Conference on Broadband, Wireless Computing, Communication and Applications  
Control Flow Flits must be transmitted so that there are no dropped flits, or with some kind of resend protocol.  ...  flow of flits.  ... 
doi:10.1109/bwcca.2010.51 dblp:conf/bwcca/MoriEAK10 fatcat:qkj6vn34xjdt7kpdg3e6qedetq

Efficient Bypass in Mesh and Torus NoCs

Iván Pérez, Enrique Vallejo, Ramón Beivide
2020 Journal of systems architecture  
NEBB-Hybrid is proposed to employ the best flow control in each situation. The mechanism is extended to torus topologies, using FBFC and shared buffers.  ...  The proposals have been evaluated using Booksim, showing up to 75% reduction of the buffered flits for single-flit packets, which translates into latency and dynamic power reductions of up to 30% and 23%  ...  Flow Control and deadlock avoidance Flow control in NoCs is applied at flit level when using WH.  ... 
doi:10.1016/j.sysarc.2020.101832 fatcat:jpio4lotvjhelfcelp5bysib7y
« Previous Showing results 1 — 15 out of 2,429 results