Soft real-time scheduling on simultaneous multithreaded processors

Jain, Rohit; Hughes, Christopher J.; Adve, Sarita V.

doi:10.1109/real.2002.1181569

Cited by 61 publications

(65 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…He did not specify the execution order, but he used an out-of-order processor as base architecture and so did most of the later SMT researchers. Consequently, most of the work on real-time and SMT is also based on out-of-order pipelines [3][4][5][6]. But the unpredictability of out-of-order pipelines does not allow hard real-time execution, only soft real-time scheduling is addressed.…”

Section: Related Workmentioning

confidence: 99%

How to Enhance a Superscalar Processor to Provide Hard Real-Time Capable In-Order SMT

Mische

Guliashvili

Uhrig

et al. 2010

Architecture of Computing Systems - ARCS 2010

View full text Add to dashboard Cite

Abstract. This paper describes how a superscalar in-order processor must be modified to support Simultaneous Multithreading (SMT) such that time-predictability is preserved for hard real-time applications. For superscalar in-order architectures the calculation of the Worst Case Execution Time (WCET) is much easier and tighter than for out-of-order architectures. By a careful enhancement that completely isolates the threads, this capability can be perpetuated to an in-order SMT architecture. Our design goal is to minimise the WCET of the highest priority thread, while releasing as many resources as possible for the execution of concurrent non critical threads. The resultant processor executes hard real-time threads at the same speed as its singlethreaded ancestor, but idle issue slots are dynamically used by non critical threads. The modifications to enable SMT include a multithreaded fetch stage, an additional real-time issue stage, a wider register set, a prioritised multithreaded memory interface, split phase loads and interruptible microcodes for multi-cycle operations. The application of these enhancements is demonstrated by CarCore, a multithreaded embedded processor that implements the Infineon Tricore instruction set.

show abstract

Section: Related Workmentioning

confidence: 99%

How to Enhance a Superscalar Processor to Provide Hard Real-Time Capable In-Order SMT

Mische

Guliashvili

Uhrig

et al. 2010

Architecture of Computing Systems - ARCS 2010

View full text Add to dashboard Cite

show abstract

“…Peha et al [2] suggested some new approaches towards scheduling but they are not QoS-aware. Matschulat et al [3] and Tomoyoshi et al [4] have studied QoS [5][6][7] in embedded systems but none of them considered Markov chains. The concept of system QoS based on Markov statistics, is for the first time we have worked out here to our best knowledge.…”

Section: Related Workmentioning

confidence: 99%

A Novel Embedded Statistical Scheduler for Realising High System-QoS

Paul¹,

Jana²,

Pandit³

2010

IJCSIT

View full text Add to dashboard Cite

ABSTRACT-Development of schedulers for real-time embedded systems is an emerging area of research as it addresses timely scheduling issues like co-operative scheduling KEYWORDS-QoS, Schedulers, Embedded system, Markov proces , Probability of Error INTRODUTIONOptimizing the CPU usage times allocated to processes for providing high QoS is a key issue and challenge in modern embedded system design. In a multi-tasking environment where an embedded processor runs several processes, the CPU usage time provisioned for each process is analytically intractable and needs simulation study. In our simulation we consider Markov statistics to get the initial estimates for the scheduled times. They are later dynamically tuned by the scheduler using feedback control for maintaining minimum aggregate probability of error of the system and thus best system quality of service (QoS). RELATED WORKCazorla et al [1] have studied QoS for embedded systems. But the QoS considered is specific for an individual process. Peha et al [2] suggested some new approaches towards scheduling but they are not QoS-aware. Matschulat et al [3] and Tomoyoshi et al [4] have studied QoS [5][6][7] in embedded systems but none of them considered Markov chains. The concept of system QoS based on Markov statistics, is for the first time we have worked out here to our best knowledge.Most of the earlier work [3-6] on finite state modelling of schedulers considered the states of each process as of four types: 1) sleeping 2) ready 3) executing and 4) blocked. This is not an enabling technique to calculate system QoS. We, on the other hand, model here each of the processes in a multi-tasking environment as an individual Markov process and calculate the system QoS. To our best of knowledge, this is the first such approach [8,9]. METHOLODGYThis paper reports a Finite-state machine (FSM) based Markov chain model for schedulers where each process is modelled as a Markov state. The processes settle to a steady state

show abstract

“…Real-time scheduling by software on an SMT processor is detailed in [8]. This research focuses on soft real-time scheduling for simultaneous execution threads and their shared resources on the SMT processor.…”

Section: Related Workmentioning

confidence: 99%

“…The RMT architecture accurately guarantees the execution time of the prioritized threads. Software scheduling such as in [8] may improve execution time precision of lower priority threads.…”

Section: Pipelinementioning

confidence: 99%