Introduction Traditional benchmarks such as SPEC model a simple workload: a single address space, a single thread of control, no input beyond an initial data set, and little if any use of system routines. Increasingly common computer usage includes multiprogrammed environments, interactivity, and multiprocessor machines. Workloads that represent these environments have significantly different characteristics than single-process batch workloads [1]. Computer architects often rely on measurements of existing systems to drive studies of system performance. These measurements are gathered by behavior monitoring systems and might include information such as execution profiles, address traces, and instruction traces. Existing behavior monitoring...
Application-level monitoring is required for continuously operating software systems to maintain the...
One of the major architectural design considerations for any computer system is that of the memory s...
Improvements in performance and energy efficiency often require deep understanding of the complex in...
Vita.A methodology is presented by which computer empirical performance models can be obtained in a ...
Due to the complex interactions between activities in parallel processes, the dynamic behavior of th...
A generalized workload definition is presented which constructs measurable workloads of unit size fr...
Measurement of computer systems performance has existed since the earliest days of computing, but th...
This paper describes the DIGITAL Continuous Profiling Infrastructure, a sampling-based profiling sys...
Performance observability is the ability to accurately capture, analyze, and present (collectively o...
Real-time performance monitoring to identify poorly or under-performing loops has become an integral...
Understanding the behavior of current and future workloads is key for designers of future computer s...
We present accurate, low-level measurements of process preemption, interrupt handling and memory sys...
This technical report explores workload characterization using processor hardware counter sampling. ...
Performance failures are commonplace in most computing environments; without system monitoring they ...
This study develops a procedure for modeling computer systems and for analyzing their performance. H...
Application-level monitoring is required for continuously operating software systems to maintain the...
One of the major architectural design considerations for any computer system is that of the memory s...
Improvements in performance and energy efficiency often require deep understanding of the complex in...
Vita.A methodology is presented by which computer empirical performance models can be obtained in a ...
Due to the complex interactions between activities in parallel processes, the dynamic behavior of th...
A generalized workload definition is presented which constructs measurable workloads of unit size fr...
Measurement of computer systems performance has existed since the earliest days of computing, but th...
This paper describes the DIGITAL Continuous Profiling Infrastructure, a sampling-based profiling sys...
Performance observability is the ability to accurately capture, analyze, and present (collectively o...
Real-time performance monitoring to identify poorly or under-performing loops has become an integral...
Understanding the behavior of current and future workloads is key for designers of future computer s...
We present accurate, low-level measurements of process preemption, interrupt handling and memory sys...
This technical report explores workload characterization using processor hardware counter sampling. ...
Performance failures are commonplace in most computing environments; without system monitoring they ...
This study develops a procedure for modeling computer systems and for analyzing their performance. H...
Application-level monitoring is required for continuously operating software systems to maintain the...
One of the major architectural design considerations for any computer system is that of the memory s...
Improvements in performance and energy efficiency often require deep understanding of the complex in...