A Novel and Productive Network on Chip Switch for Dataflow
Abstract
Dataflow architecture
has shown its advantages in many high-performance computing cases. In dataflow
registering, a lot of information are every now and again moved among handling
components through the organization on-chip (NoC). Subsequently, the switch
configuration altogether affects the presentation of dataflow design. Normal
switches are intended for control-stream multi-center design and we discover
they are not appropriate for dataflow engineering. In this work, we examine and
separate the highlights of information moves in NoCs of dataflow engineering:
different objections, high infusion rate, and execution delicate to delay. In
light of the three highlights, we propose a novel and proficient NoC switch for
information stream design. The proposed switch upholds multi-objective;
consequently, it can move information with numerous objections in a solitary
exchange. Additionally, the switch embraces yield cushion to augment throughput
and receives non-flutter parcels to limit move delay. Trial results show that
the proposed switch can improve the presentation of dataflow engineering by
3.6x over a best in class switch.
Country : India
1 Valle Kumar Shyam
Assistant Professor, Department of Computer Science And Engineering, Malla Reddy College of Engineering for Women, Hyderabad -500100, Telangana, India
Tran A T,
Baas B M. Achieving high-performance on-chip networks with shared-buffer
routers. IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2019
Milutinovic
V, Salom J, Trifunovic N, Giorgi R. Guide to Dataflow Supercomputing (1st
edition). Springer International Publishing, 2018.
Wei L, Zhou
L. An equilibrium partitioning method for multicast traffic in 3D NoC
architecture. In Proc. The IFIP/IEEE International Conference on Very Large
Scale Integration, 2018.
Chen T S,
Du Z D, Sun N H, Wang J, Wu C Y, Chen Y J, Tema O. DianNao: A small-footprint
high-throughput accelerator for ubiquitous machine-learning In Proc. the 19th
International Conference on Architectural Support for Programming Languages and
Operating Systems., 2019.
Voitsechov
D, Etsion Y. Single-graph multiple flows: Energy efficient design alternative
for GPGPUs. In Proc. The 41st Int. Symp. Computer Architecture, 2019.
