A Scalable Analytical Memory Model for CPU Performance Prediction

Chennupati, Gopinath; Santhi, Nandakishore; Bird, Robert; Thulasidasan, Sunil; Badawy, Abdel-Hameed A.; Misra, Satyajayant; Eidenbenz, Stephan

doi:10.1007/978-3-319-72971-8_6

Cited by 9 publications

(13 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The memory traces used in most of these attempts are significant in size and time-consuming to process, thereby unscalable. However, recent attempts from Chennupati et al [22][23][24] demonstrated analytical models that scale with a small input run of a program. These attempts help predict the performance of an application on single-threaded programs.…”

Section: Reuse Distancementioning

confidence: 99%

“…We can obtain the basic blocks' labels in an LLVM intermediate representation file. We use a modified version [22] of LLVM based instrumentation tool, Byfl [62] which can instrument the preferred functions (in this case, functions starting with OUT ) to generate the basic block labeled memory trace through sequential execution. We add the basic block labels in the memory trace in a way that all the memory addresses that are accessed as a result of executing the corresponding straight-line code of (BB i ) are grouped together.…”

Section: Memory Trace Generation For Different Cache Hierarchiesmentioning

confidence: 99%

“…The conventional methods of measuring the reuse profile are costly because of the enormous size of the memory traces. The previous work of PPT used a stack-based method to compute the reuse profiles of a program [5,6,22]. The stack-based method has a worst-case time complexity of O(N.M ) for a trace of length N containing M distinct references.…”

Section: Reuse Profile Calculationmentioning

confidence: 99%

“…We measure private, and shared cache hit rates using an analytical memory model (SDCM) proposed by Brehob and Enbody [17]. SDCM was used before to predict the cache hit rates of CPUs [11,[22][23][24] and GPUs [5,6] in PPT. Equation 1 shows the way to measure the conditional hit rate at a given reuse distance (P (h | D)).…”

Section: Hit Rate Estimationmentioning

confidence: 99%

See 3 more Smart Citations

PPT-Multicore: Performance Prediction of OpenMP applications using Reuse Profiles and Analytical Modeling

Barai,

Arafa,

Badawy

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

We present PPT-Multicore, an analytical model embedded in the Performance Prediction Toolkit (PPT) to predict parallel application performance running on a multicore processor. PPT-Multicore builds upon our previous work towards a multicore cache model. We extract LLVM basic block labeled memory trace using an architecture-independent LLVM-based instrumentation tool only once in an application's lifetime. The model uses the memory trace and other parameters from an instrumented sequentially executed binary. We use a probabilistic and computationally efficient reuse profile to predict the cache hit rates and runtimes of OpenMP programs' parallel sections. We model Intel's Broadwell, Haswell, and AMD's Zen2 architectures and validate our framework using different applications from PolyBench and PARSEC benchmark suites. The results show that PPT-Multicore can predict cache hit rates with an overall average error rate of 1.23% while predicting the runtime with an error rate of 9.08%.

show abstract

Section: Reuse Distancementioning

confidence: 99%

Section: Memory Trace Generation For Different Cache Hierarchiesmentioning

confidence: 99%

Section: Reuse Profile Calculationmentioning

confidence: 99%

Section: Hit Rate Estimationmentioning

confidence: 99%

See 2 more Smart Citations

PPT-Multicore: Performance Prediction of OpenMP applications using Reuse Profiles and Analytical Modeling

Barai,

Arafa,

Badawy

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…The above process can be time-consuming when we desire to predict for a range of parameters. Some analytical models [13,14,21,22,34] require the conversion of source code into a control flow chart for the ease of framing equations. Our model predicts the performance of an application on a range of input parameters without requiring a new set of equations, as FiM uses a machine-learning model to learn hardware parameters.…”

Section: Related Workmentioning

confidence: 99%

New Performance Modeling Methods for Parallel Data Processing Applications

Bhimani

Leeser

et al. 2019

ACM Trans. Model. Comput. Simul.

View full text Add to dashboard Cite

Predicting the performance of an application running on parallel computing platforms is increasingly becoming important because of its influence on development time and resource management. However, predicting the performance with respect to parallel processes is complex for iterative and multi-stage applications. This research proposes a performance approximation approach FiM to predict the calculation time with FiM-Cal and communication time with FiM-Com of an application running on a distributed framework. FiM-Cal consists of two key components that are coupled with each other: (1) a Stochastic Markov Model to capture nondeterministic runtime that often depends on parallel resources, e.g., number of processes, and (2) a machinelearning model that extrapolates the parameters for calibrating our Markov model when we have changes in application parameters such as dataset. Along with the parallel calculation time, parallel computing platforms consume some data transfer time to communicate among different nodes. FiM-Com consists of a simulation queuing model to quickly estimate communication time. Our new modeling approach considers different design choices along multiple dimensions, namely (i) process-level parallelism, (ii) distribution of cores on multi-processor platform, (iii) application related parameters, and (iv) characteristics of datasets. The major contribution of our prediction approach is that FiM can provide an accurate prediction of parallel processing time for the datasets that have a much larger size than that of the training datasets. We evaluate our approach with NAS Parallel Benchmarks and real iterative data processing applications. We compare the predicted results (e.g., end-to-end execution time) with actual experimental measurements on a real distributed platform. We also compare our work with an existing prediction technique based on machine learning. We rank the number of processes according to the actual and predicted results from FiM and calculate the correlation between the actual and predicted rankings. Our results show that FiM obtains a high correlation in the range of 0.80 to 0.99, which indicates considerable accuracy of our technique. Such prediction provides data analysts a useful insight of optimal configuration of parallel resources (e.g., number of processes and number of cores) and also helps system designers to investigate the impact of changes in application parameters on system performance.

show abstract