1. |
- Chen, Yancang, et al.
(author)
-
A Trace-driven Hardware-level Simulator for Design and Verification of Network-on-Chips
- 2010
-
In: 2011 INTERNATIONAL CONFERENCE ON COMPUTERS, COMMUNICATIONS, CONTROL AND AUTOMATION (CCCA 2011), VOL II. - : IEEE. ; , s. 32-35
-
Conference paper (peer-reviewed)abstract
- Traditional communications of general-purpose multi-core processor and application-specific System-on-Chip face challenges in terms of scalability and complexity. Network-on-Chip (NoC) has been the most promising solution for the communications of multi-core and many-core chips. In this paper, we present a trace-driven hardware-level simulator (noted HS) based on SystemVerilog for the design and verification of NoCs. Different from the state-of-the-art NoC simulators, the HS owns three important characteristics in addition to the capability of creating simulation and synthesizable NoC descriptions: 1) hardware-level simulation can be done, which means more implementation details of hardware than flit-level simulation; 2) router debugging and verification can be done at RTL by inserting assertions and coverage; 3) trace-based application simulations can be done besides synthetic workloads. A 4 X 4 2D mesh NoC with output virtual-channel routers verifies the capability of our HS.
|
|
2. |
- Chen, Yancang, et al.
(author)
-
A single-cycle output buffered router with layered switching for Networks-on-Chips
- 2012
-
In: Computers & electrical engineering. - : Elsevier BV. - 0045-7906 .- 1879-0755. ; 38:4, s. 906-916
-
Journal article (peer-reviewed)abstract
- We present a single-cycle output buffered router based on layered switching for networks on chips (NoCs). Different from state-of-the-art NoC routers, the router has three important characteristics: (1) It employs layered switching, which implements wormhole on top of virtual cut-through (VCT) switching; (2) In contrast to input buffered architectures, it adopts an output buffered architecture; (3) It is single cycle, meaning that the router pipeline takes only one cycle for all flits. Experimental results show that the router achieves up to 80% of ideal network throughput under uniform random traffic pattern. Compared with wormhole switching, layered switching achieves up to 36.9% latency reduction for 12-flit packets under uniform random traffic with an injection rate of 0.5 flit/cycle/node. Under 65 nm technology synthesized results show that its critical path has only 20 logic gates, and it reduces 11% area compared to the input virtual-channel router with the same buffer capacity.
|
|