Online resource management in a multiprocessor with a network-on-chip

Abstract

We propose an online resource allocation solution for multiprocessor systems-on-chip, that executes several real-time, streaming media jobs simultaneously. The system consists of up to 24 processors connected by an AEthereal [7] Network-on-Chip (NoC) of 4 to 12 routers. A job is a set of processing tasks connected by FIFO channels. Each job can be independently started or stopped by the user. Each job is annotated with resource budgets per computation task and communication channel which have been computed at compile-time. When a job is requested to start, resources that meet the required resource budgets have to be found. Because it is done online, allocation must be done with low-complexity algorithms. We do the allocation in two-steps. First, tasks are assigned to virtual tiles (VTs), while trying to minimise the total number of VTs and the total bandwidth used. In the second step, these VTs are mapped to real tiles, and network bandwidth allocation and routing are performed simultaneously. We show with simulations that introducing randomisation in the processing order yields a significant improvement in the percentage of mapping succdesses. In combination, these techniques allow 95% of the processor resources to be allocated while handling a large number of job arrivals and departures.

Buffer Sizing for Rate-Optimal Single-Rate Data-Flow Scheduling Revisited

“…The following definition of a buffer-capacity distribution function differs from the one given in Equation (5) in that it assigns a separate buffer to each edge, and that it reserves space in each buffer for an active firing of the consuming actor. This definition captures the most conservative buffering requirements that are possible.…”

“…In a rate-optimal schedule, which has period µ, the cycle through A, A j , and P Aj enforces that the firings of these three actors occur as soon as the firing of the preceding actor finishes. As a consequence, the buffers of A and P Aj in G O are 0 and (due to the initial token) 1 respectively 5 . As another consequence, the firing of A j in G O completes at the same point in time as the corresponding firing of A completes in G g .…”

“…This allows to verify real-time requirements. It is furthermore possible to use the abstract concepts of actors and edges to model properties of communication channels, memory mapping, and scheduler settings [3]- [5], therefore allowing for the analysis of dataflow graph implementations onto specific architectures.…”

Run-time Spatial Mapping of Streaming Applications to Heterogeneous Multi-Processor Systems

“…Their run-time spatial mapping approaches are then related to how well their results are improved by the designtime preparations [31,32] or how well (assumed) detrimental factors are avoided [17]. The work presented in [12] only considers independent tasks, i.e.…”

“…In [13,17], some conditions that must be met by resource management policies for real-time applications are described. The conditions mentioned are admission control (applications are only allowed to start if sufficient resources can be allocated) and guaranteed resource provisions (running tasks are always allowed access to the resources allocated to them).…”