The integration of many processing elements per die makes it more difficult to provide low latency in the Network-on-Chip (NoC). Multihop bypass proposals, such as SMART, attack this problem by allowing flits to skip multiple routers in the path in a single cycle, drastically reducing latency while preserving a regular tiled layout. However, multihop bypass routers are more complex and relatively different from traditional NoC routers, since they rely on global broadcast signals and global allocation mechanisms. Additionally, the maximum number of nodes that can be bypassed within a single cycle is limited by the Critical Path Delay (CPD) of the NoC. Hence, a practical multihop bypass mechanism must also minimize this delay.

Many-core processors demand scalable, efficient and low latency NoCs. Bypass routers are an affordable solution to attain low latency in relatively simple topologies like the mesh. SMART improves on traditional bypass routers implementing multi-hop bypass which reduces the importance of the distance between pairs of nodes. Nevertheless, the conservative buffer reallocation policy of SMART requires a large number of Virtual Channels (VCs) to offer high performance, penalizing its implementation cost. Besides, SMART zero-load latency values highly depend on HPCMaxHPCMax, the maximum number of hops that can be jumped per cycle. In this article, we present Speculative-SMART++ (S-SMART++), with two mechanisms that significantly improve multi-hop bypass. First, zero-load latency is reduced by speculatively setting consecutive multi-hops. Second, the inefficient buffer reallocation policy of SMART is reduced by combining multi-packet buffers, Non-Empty Buffer Bypass and per-packet allocation. These proposals are evaluated using functional simulation, with synthetic and real loads, and synthesis tools. S-SMART++ does not need VCs to obtain the performance of SMART with 8 VCs, reducing notably logic resources and dynamic power. Additionally, S-SMART++ reduces the base-latency of SMART by at least 29.2 percent, even when using the biggest HPCMaxHPCMaxpossible.

Network-on-Chips are a critical part of modernmultiprocessors and their relevance will grow with the number ofcores. The development of future NoC designs relies on detailedsimulation models that accurately estimate their performance, power and hardware cost. Bypass routers are very relevant and promising proposals dueto their improved performance. Bypass routers reduce latencythanks to a combination of speculation, pre-routing (lookaheadrouting) and buffer bypass, which also reduce energy consumption by avoiding unnecessary buffer writes and reads. Multi-hop bypass NoCs, known as SMART, even bypass the crossbar of multiple routers in a single cycle. However, publicly available NoC simulators, such as BookSim or Garnet, do not implement bypass mechanisms or do not model them accurately. In this work, we present Bypass Simulation Toolset(BST), a set of tools to accurately simulate NoCs with single-and multi-hop bypass routers. BST combines and extends several simulation tools: an extension of BookSim with state-of-the-art cycle-accurate bypass router models and additional flow control mechanisms; an RTL implementation of multi-hop bypass mechanisms based on OpenSMART; an API to ease a modular integration of the BST NoC simulator in full system simulators; and a set of scripts to automate simulation execution and data collection. To showcase BST, we i) validate BookSim SMART models with the RTL implementation; ii) compare bypass and traditional non-bypass router models; iii) integrate BookSim in gem5 using the proposed API and compare it with gem5's Simple and Garnet 2.0 NoC models; and iv) present a case study evaluating different combinations of router types and topologies recently proposed for NoCs, highlighting the flexibility of the BST toolset. The toolset is available at www.atc.unican.es/software.html.

Low latency and low implementation cost are two key requirements in NoCs. SMART routers implement multi-hop bypass, obtaining latency values close to an ideal point-to-point interconnect. However, it requires a significant amount of resources such as Virtual Channels (VCs), which are not used as efficiently as possible, preventing bypass in certain scenarios. This translates into increased area and delay, compared to an ideal implementation.