Zobrazeno 1 - 10
of 192
pro vyhledávání: '"Bisection bandwidth"'
Publikováno v:
IEEE Transactions on Parallel and Distributed Systems. 33:1054-1068
Interconnection networks play an essential role in the architecture of high-performance computing (HPC) systems. In this article, we explore the Galaxyfly family to build flexible-scale interconnection networks. Galaxyfly is guaranteed to retain a sm
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::ea5bb7ea81603069f8ffa433943e47e5
https://doi.org/10.1109/tpds.2021.3100783
https://doi.org/10.1109/tpds.2021.3100783
Publikováno v:
IEEE Transactions on Communications. 69:8363-8374
Data center networks typically adopt multi-rooted tree topologies to provide high bisection bandwidth. Various fine-grained load balancing schemes have been proposed to split flows across multiple paths. However, data center networks suffer from many
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::994958be7334daa89aa693575a161d97
https://doi.org/10.1109/tcomm.2021.3118467
https://doi.org/10.1109/tcomm.2021.3118467
Autor:
Weihe Li, Lyu Wenjun, Wenchao Jiang, Tian He, Jianxin Wang, Jiawei Huang, Zhaoyi Li, Jinbin Hu
Publikováno v:
IEEE/ACM Transactions on Networking. 29:2367-2384
The state-of-the-art datacenter load balancing designs commonly optimize bisection bandwidth with homogeneous switching granularity. Their performances surprisingly degrade under mixed traffic containing both short and long flows. Specifically, the s
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::af006488bae1e2f108319865e5338ef3
https://doi.org/10.1109/tnet.2021.3088276
https://doi.org/10.1109/tnet.2021.3088276
Publikováno v:
Journal of Cloud Computing: Advances, Systems and Applications, Vol 10, Iss 1, Pp 1-16 (2021)
Modern data center networks typically adopt multi-rooted tree topologies such leaf-spine and fat-tree to provide high bisection bandwidth. Load balancing is critical to achieve low latency and high throughput. Although the per-packet schemes such as
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::2888e7a5f11fc629eeb89d22b0c308d5
https://doaj.org/article/7c963494cf5649588d0319fb64e6c525
https://doaj.org/article/7c963494cf5649588d0319fb64e6c525
Publikováno v:
China Communications. 18:198-212
Cloud providers (e.g., Google, Alibaba, Amazon) own large-scale datacenter networks that comprise thousands of switches and links. A load-balancing mechanism is supposed to effectively utilize the bisection bandwidth. Both Equal-Cost Multi-Path (ECMP
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::1e0be1d81d4ac328f6235b74bb4cfd8d
https://doi.org/10.23919/jcc.2021.04.015
https://doi.org/10.23919/jcc.2021.04.015
Publikováno v:
IEEE Access, Vol 9, Pp 35539-35548 (2021)
Modern Data Center Networks (DCNs) are commonly based on Clos topologies with a large number of equal-cost multiple paths to provide high bisection bandwidth. The existing Random Packet Spraying (RPS) scheme spreads each flow of packets to all availa
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::0a239cb51974e16e8577d3346da03d7a
https://doi.org/10.1109/access.2021.3062866
https://doi.org/10.1109/access.2021.3062866
Autor:
Luis A. Plana, Steve Temple, Jeffrey Pepper, Jonathan Heathcote, Simon Davidson, Jim Garside, Steve Furber, Mikel Luján
Publikováno v:
IEEE Access, Vol 8, Pp 84918-84928 (2020)
Plana, L A, Garside, J, Heathcote, J, Pepper, J, Temple, S, Davidson, S, Luján, M & Furber, S 2020, ' spiNNlink: FPGA-Based Interconnect for the Million-Core SpiNNaker System ', IEEE Access, vol. 8, 9079810, pp. 84918-84928 . https://doi.org/10.1109/ACCESS.2020.2991038
IEEE Access
Plana, L A, Garside, J, Heathcote, J, Pepper, J, Temple, S, Davidson, S, Luján, M & Furber, S 2020, ' spiNNlink: FPGA-Based Interconnect for the Million-Core SpiNNaker System ', IEEE Access, vol. 8, 9079810, pp. 84918-84928 . https://doi.org/10.1109/ACCESS.2020.2991038
IEEE Access
SpiNNaker is a massively-parallel computer system optimised for the simulation, in real time, of very large networks of spiking neurons. The system consists of over 1 million, energy-efficient ARM cores distributed over 57,600 SpiNNaker chips, each o
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::c854158f41f383f6557881d485711e72
https://ieeexplore.ieee.org/document/9079810/
https://ieeexplore.ieee.org/document/9079810/
Publikováno v:
SIGCOMM
While prior work has explored many proposed datacenter designs, only two designs, Clos-based and expander-based, are generally considered practical because they can scale using commodity switching chips. Prior work has used two different metrics, bis
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::b066d7e30ad3ed03c78139096e1f1b24
https://doi.org/10.1145/3452296.3472913
https://doi.org/10.1145/3452296.3472913
Publikováno v:
NetSoft
Data center networks are designed with multi-rooted topologies to provide the large bisection bandwidth. These topologies provide high path diversity and need a load balancing algorithm to utilize bisection bandwidth efficiently. To distributes traff
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::d2c438e20f80c80213071b73eac640c1
https://doi.org/10.1109/netsoft51509.2021.9492566
https://doi.org/10.1109/netsoft51509.2021.9492566
Publikováno v:
SIGMOD Conference
The recent scale-up of GPU hardware through the integration of multiple GPUs into a single machine and the introduction of higher bandwidth interconnects like NVLink 2.0 has enabled new opportunities of relational query processing on multiple GPUs. H
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::3657b35efce4f269468a9a5586082f97
https://doi.org/10.1145/3448016.3457254
https://doi.org/10.1145/3448016.3457254