This thesis investigates the capabilities of Palladio to predict the performance of software/hardware systems. The Palladio simulations are accurate for systems which run on single core processors. Experiments showed that the predictions are not accurate for multicore systems. The parallelization of programs is complex. In addition a parallelized program executed on four cores is not automatically four times faster than the single core program. There are reasons for this on the software/code side (e.g. Amdahl's law) but also on the hardware side (e.g. memory bandwidth). The so called memory bandwidth is referring to the capacity limit of the memory bus, the bus from the CPU to the memory. The memory bandwidth is theoretically becoming a mor...
Parallelism is ubiquitous in modern computer architectures. Heterogeneity of CPU cores and deep memo...
The CPUs, memory, interconnection network, operating system, runtime system, I/O subsystem, and appl...
peer-reviewedThe shift towards multicore processing has led to a much wider population of developer...
Model-based performance prediction is a well-known concept to ensure the quality of software.Current...
Die modellbasierte Performancevorhersage ist ein bekanntes Konzept zur Gewährleistung der Softwarequ...
For the last two decades, the number of cores in modern CPUs has been steadily increasing. This enab...
Shared memory architectures are widely taking place. Following the structured parallel programming a...
Performance analysis is the task of monitor the behavior of a program execution. The main goal is to...
International audienceTo amortize the cost of MPI communications, distributed parallel HPC applicati...
Most computing systems are heavily dependent on their main memories, as their primary storage, or as...
To increase performance, modern processors employ complex techniques such as out-of-order pipelines ...
This thesis presents a unified approach to modeling of parallel architectures and algorithms with sp...
We propose a new model with small degreee of parallelism that reflects current and future multicore ...
As the complexity of parallel computers grows, constraints posed by the construction of larger syste...
Since many years, we observe a shift from classical multiprocessor systems tomulticores, which tight...
Parallelism is ubiquitous in modern computer architectures. Heterogeneity of CPU cores and deep memo...
The CPUs, memory, interconnection network, operating system, runtime system, I/O subsystem, and appl...
peer-reviewedThe shift towards multicore processing has led to a much wider population of developer...
Model-based performance prediction is a well-known concept to ensure the quality of software.Current...
Die modellbasierte Performancevorhersage ist ein bekanntes Konzept zur Gewährleistung der Softwarequ...
For the last two decades, the number of cores in modern CPUs has been steadily increasing. This enab...
Shared memory architectures are widely taking place. Following the structured parallel programming a...
Performance analysis is the task of monitor the behavior of a program execution. The main goal is to...
International audienceTo amortize the cost of MPI communications, distributed parallel HPC applicati...
Most computing systems are heavily dependent on their main memories, as their primary storage, or as...
To increase performance, modern processors employ complex techniques such as out-of-order pipelines ...
This thesis presents a unified approach to modeling of parallel architectures and algorithms with sp...
We propose a new model with small degreee of parallelism that reflects current and future multicore ...
As the complexity of parallel computers grows, constraints posed by the construction of larger syste...
Since many years, we observe a shift from classical multiprocessor systems tomulticores, which tight...
Parallelism is ubiquitous in modern computer architectures. Heterogeneity of CPU cores and deep memo...
The CPUs, memory, interconnection network, operating system, runtime system, I/O subsystem, and appl...
peer-reviewedThe shift towards multicore processing has led to a much wider population of developer...