1,721,017 research outputs found
CPU power estimation using PMCs and its application in gem5
No full textFast and accurate estimation of CPU power consumption is necessary to inform run-time power management approaches and allow effective design space exploration. Power simulators, combined with a full-system architectural simulator such as gem5, enable power-performance trade-offs to be investigated early in the design of a system. However, the accuracy of existing power simulators is known to be low, and this can lead to incorrect conclusions being made. In this talk, I will present our statistically rigorous methodology for building accurate run-time power models using Performance Monitoring Counters (PMCs) for mobile and embedded devices, and demonstrate how our models make more efficient use of limited training data and better adapt to unseen scenarios by uniquely considering stability. Models built using the methodology for both ARM Cortex-A7 and Cortex-A15 CPUs exhibit a 3.8% and 2.8% average error respectively. I will also present online resources that we have made available from the work, including software tools, documentation, raw data and further results. I will also present results from an investigation into the correlation between gem5 activity statistics and hardware PMCs. Based on this, a gem5 power model for a simulated quadcore ARM Cortex-A15 has been created, built using the above methodology, and its accuracy compared against experimental results obtained from hardware
Dataset supporting the article entitled "Energy Harvesting and Transient Computing: A Paradigm Shift for Embedded Systems?"
No full textThis dataset supports the article entitled "Energy Harvesting and Transient Computing: A Paradigm Shift for Embedded Systems?" accepted as invited paper for DAC 2016.
</span
Power-neutral computing for IoT devices
No full textEnergy harvesting has shown considerable promise for IoT systems, particularly in applications where battery replacement or recharging is undesirable. However, energy harvesting (often referred to as 'energy-neutral') systems, typically add energy storage (a battery or supercapacitor) to smooth temporal dynamics in source. For many applications where space is premium (e.g. wearable/implantable) this is problematic. In this presentation, I will present the concept of 'power-neutral' computing, where systems operate from the instantaneous available power without requiring energy storage to be added to the system. In particular, I will present a control approach to adapt the power consumption on IoT microcontrollers which uses software-based maximum power point tracking (MPPT) to maximise the extraction and consumption of power from the energy harvesting supply. It does this without any additional hardware being added to the platform, and is the first MPPT approach we are aware of that performs this using only software. The MPPT approach achieves this by leveraging the opportunities provided by power-neutral computing, and operates by adaptively adjusting properties of the control algorithm. Practical validation of the power-neutral approach demonstrates successful operation without added energy storage, and allows execution of nearly 50% more instructions compared to static approaches
Dataset supporting the article entitled "Energy-Driven Computing: Rethinking the Design of Energy Harvesting Systems"
No full textThis dataset supports the article entitled "Energy-Driven Computing: Rethinking the Design of Energy Harvesting Systems" accepted for publication in DATE 2017.</span
Dataset for 'ANTS'21 W5 AIMLECV - Route Lifetime Analysis in Vehicular Networks'
No full textThis is the dataset of the accepted paper (Nov, 2021): T. Ivanescu, H. Yetgin, M. El-Hajjar, G. Merrett, "Route Lifetime Analysis for Vehicular Networks". </span
Dataset supporting the article entitled "The Slowdown or Race-to-idle Question: Workload-Aware Energy Optimization of SMT Multicore Platforms under Process Variation"
No full textThis dataset supports the article entitled "The Slowdown or Race-to-idle Question: Workload-Aware Energy Optimization of SMT Multicore Platforms under Process Variation" accepted for publication in DATE conference 2016.</span
Dataset supporting the paper entitled “Power neutral performance scaling for energy harvesting MP-SoCs”
No full textThis dataset supports the paper entitled “Power Neutral Performance Scaling for Energy Harvesting MP-SoCs” accepted for publication</span
Dataset for "Exploring the Effect of Energy Storage Sizing on Intermittent Computing System Performance"
No full textThis dataset supports the article entitled "Exploring the Effect of Energy Storage Sizing on Intermittent Computing System Performance" accepted for publication in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.</span
Dataset for "GhostShiftAddNet: More Features from Energy-Efficient Operations"
No full textThis dataset supports the publication: GhostShiftAddNet: More Features from Energy-Efficient Operations.' in 'British Machine Vision Conference 2021'.</span
Dataset for PhD Thesis "Energy Budgeting for Intermittently-Powered Systems"
No full textThis is a Dataset to support the Southampton Doctoral thesis "Energy Budgeting for Intermittently-Powered Systems"
This dataset contains:
'variable_data.csv': Data supporting Figure 5.1 \Delta V_task varying linearly with the data size in AES 128-bit encryption.
'pv_curve.csv': Data supporting Figure 5.2 An I-V curve of a glass-type amorphous PV panel (Sanyo AM-1417CA, 35mm x 13.9mm) under a white LED lighting condition.
'perf.csv': Data supporting Figure 5.3 Numbers of completed and failed operations of DEBS Low, Samoyed, DEBS High, and OPTIC Oracle given random data sizes and configurations and a PV supply in a 10s simulation.
'cap.csv': Data supporting Figure 5.5 Number of completed operations of Samoyed, DEBS High, and the OPTIC Oracle with capacitance reduction.
'profiling_accuracy.csv': Data supporting Figure 5.11 Error Distribution of OPTIC's Runtime Energy Profiling given a PV supply, compared to the "Naive" disconnecting-supply method.
'analog.csv', 'digital.csv': Data supporting Figure 5.12 A voltage trace of OPTIC adapting to a new operation on a new device.
'cap_test_aes.csv', 'cap_test_dma.csv', 'cap_test_radio.csv': Data supporting Figure 5.13 Effect of Capacitor Degradation on OPTIC and DEBS.
'datasize.csv': Data supporting Figure 5.14 Relative Completion Rates of Samoyed, DEBS, and OPTIC with variable data sizes and a PV supply.
Dataset for Chapter 3 and Chapter 4 of the thesis is already published. DOI: 10.5258/SOTON/D1785
</span
- …
