Publications Workload Profiling
E-WarP: a System-wide Framework for Memory Bandwidth Profiling and Management
41st IEEE Real-Time Systems Symposium (RTSS 2020), December 2020, Houston, TX, USA
Best Student Paper Award
Abstract
The proliferation of multi-core, accelerator-enabled embedded systems has introduced new opportunities to consolidate real-time systems of increasing complexity. But the road to build confidence on the temporal behavior of co-running applications has presented formidable challenges. Most prominently, the main memory subsystem represents a performance bottleneck for both CPUs and accelerators. And industry-viable frameworks for full-system main memory management and performance analysis are past due. In this paper, we propose our Envelope-aWare Predictive model, or E-WarP for short. E-WarP is a methodology and technological framework to: (1) analyze the memory demand of applications following a profile-driven approach; (2) make realistic predictions on the temporal behavior of workload deployed on CPUs and accelerators; and (3) perform saturation-aware system consolidation. This work aims at providing the technological foundations as well as the theoretical grassroots for truly workload-aware analysis of real-time systems. We provide a full implementation of our techniques on a commercial platform (NXP S32V234) and make two key observations. First, we achieve, on average, a 6% over-prediction on the runtime of bandwidth-regulated applications. Second, we experimentally validate that the calculated bounds hold if the main memory subsystem operates below saturation.
Code and hardware artifacts
Cite this paper
@inproceedings{ewarp20,
title = {{E-WarP: a System-wide Framework for Memory Bandwidth Profiling and Management}},
author = {Sohal, Parul and Tabish, Rohan and Drepper, Ulrich and Mancuso, Renato},
booktitle = {41st IEEE Real-Time Systems Symposium (RTSS 2020)},
year = 2020,
month = dec,
pages = {345--357},
address = {Houston, TX, USA},
doi = {10.1109/RTSS49844.2020.00039},
url = {https://doi.org/10.1109/RTSS49844.2020.00039},
note = {Best Student Paper Award}
}
TY - CONF AU - Sohal, Parul AU - Tabish, Rohan AU - Drepper, Ulrich AU - Mancuso, Renato TI - E-WarP: a System-wide Framework for Memory Bandwidth Profiling and Management T2 - 41st IEEE Real-Time Systems Symposium (RTSS 2020) PY - 2020 DA - 2020/12// CY - Houston, TX, USA SP - 345 EP - 357 DO - 10.1109/RTSS49844.2020.00039 UR - https://doi.org/10.1109/RTSS49844.2020.00039 AB - The proliferation of multi-core, accelerator-enabled embedded systems has introduced new opportunities to consolidate real-time systems of increasing complexity. But the road to build confidence on the temporal behavior of co-running applications has presented formidable challenges. Most prominently, the main memory subsystem represents a performance bottleneck for both CPUs and accelerators. And industry-viable frameworks for full-system main memory management and performance analysis are past due. In this paper, we propose our Envelope-aWare Predictive model, or E-WarP for short. E-WarP is a methodology and technological framework to: (1) analyze the memory demand of applications following a profile-driven approach; (2) make realistic predictions on the temporal behavior of workload deployed on CPUs and accelerators; and (3) perform saturation-aware system consolidation. This work aims at providing the technological foundations as well as the theoretical grassroots for truly workload-aware analysis of real-time systems. We provide a full implementation of our techniques on a commercial platform (NXP S32V234) and make two key observations. First, we achieve, on average, a 6% over-prediction on the runtime of bandwidth-regulated applications. Second, we experimentally validate that the calculated bounds hold if the main memory subsystem operates below saturation. ER -
%0 Conference Paper %A Sohal, Parul %A Tabish, Rohan %A Drepper, Ulrich %A Mancuso, Renato %T E-WarP: a System-wide Framework for Memory Bandwidth Profiling and Management %B 41st IEEE Real-Time Systems Symposium (RTSS 2020) %D 2020 %P 345-357 %C Houston, TX, USA %R 10.1109/RTSS49844.2020.00039 %U https://doi.org/10.1109/RTSS49844.2020.00039 %X The proliferation of multi-core, accelerator-enabled embedded systems has introduced new opportunities to consolidate real-time systems of increasing complexity. But the road to build confidence on the temporal behavior of co-running applications has presented formidable challenges. Most prominently, the main memory subsystem represents a performance bottleneck for both CPUs and accelerators. And industry-viable frameworks for full-system main memory management and performance analysis are past due. In this paper, we propose our Envelope-aWare Predictive model, or E-WarP for short. E-WarP is a methodology and technological framework to: (1) analyze the memory demand of applications following a profile-driven approach; (2) make realistic predictions on the temporal behavior of workload deployed on CPUs and accelerators; and (3) perform saturation-aware system consolidation. This work aims at providing the technological foundations as well as the theoretical grassroots for truly workload-aware analysis of real-time systems. We provide a full implementation of our techniques on a commercial platform (NXP S32V234) and make two key observations. First, we achieve, on average, a 6% over-prediction on the runtime of bandwidth-regulated applications. Second, we experimentally validate that the calculated bounds hold if the main memory subsystem operates below saturation.
[
{
"id": "ewarp20",
"type": "paper-conference",
"title": "E-WarP: a System-wide Framework for Memory Bandwidth Profiling and Management",
"author": [
{
"family": "Sohal",
"given": "Parul"
},
{
"family": "Tabish",
"given": "Rohan"
},
{
"family": "Drepper",
"given": "Ulrich"
},
{
"family": "Mancuso",
"given": "Renato"
}
],
"container-title": "41st IEEE Real-Time Systems Symposium (RTSS 2020)",
"issued": {
"date-parts": [
[
2020,
12
]
]
},
"page": "345-357",
"publisher-place": "Houston, TX, USA",
"DOI": "10.1109/RTSS49844.2020.00039",
"abstract": "The proliferation of multi-core, accelerator-enabled embedded systems has introduced new opportunities to consolidate real-time systems of increasing complexity. But the road to build confidence on the temporal behavior of co-running applications has presented formidable challenges. Most prominently, the main memory subsystem represents a performance bottleneck for both CPUs and accelerators. And industry-viable frameworks for full-system main memory management and performance analysis are past due. In this paper, we propose our Envelope-aWare Predictive model, or E-WarP for short. E-WarP is a methodology and technological framework to: (1) analyze the memory demand of applications following a profile-driven approach; (2) make realistic predictions on the temporal behavior of workload deployed on CPUs and accelerators; and (3) perform saturation-aware system consolidation. This work aims at providing the technological foundations as well as the theoretical grassroots for truly workload-aware analysis of real-time systems. We provide a full implementation of our techniques on a commercial platform (NXP S32V234) and make two key observations. First, we achieve, on average, a 6% over-prediction on the runtime of bandwidth-regulated applications. Second, we experimentally validate that the calculated bounds hold if the main memory subsystem operates below saturation."
}
]
P. Sohal, R. Tabish, U. Drepper, and R. Mancuso, “E-WarP: a System-wide Framework for Memory Bandwidth Profiling and Management,” in 41st IEEE Real-Time Systems Symposium (RTSS 2020), pp. 345–357, Dec. 2020, doi: 10.1109/RTSS49844.2020.00039.
Sohal, P., Tabish, R., Drepper, U., & Mancuso, R. (2020). E-WarP: a System-wide Framework for Memory Bandwidth Profiling and Management. In 41st IEEE Real-Time Systems Symposium (RTSS 2020) (pp. 345–357). https://doi.org/10.1109/RTSS49844.2020.00039