EDEN: Enabling energy-efficient, high-performance deep neural network inference using approximate DRAM

S Koppula, L Orosa, AG Yağlıkçı, R Azizi… - Proceedings of the …, 2019‏ - dl.acm.org
The effectiveness of deep neural networks (DNN) in vision, speech, and language
processing has prompted a tremendous demand for energy-efficient high-performance DNN …

DRHEFT: Deadline-constrained reliability-aware HEFT algorithm for real-time heterogeneous MPSoC systems

J Zhou, M Zhang, J Sun, T Wang… - IEEE Transactions on …, 2020‏ - ieeexplore.ieee.org
Heterogeneous multiprocessor system-on-chips (MPSoCs) are suitable platforms for real-
time embedded applications that require powerful parallel processing capability as well as …

Approximate computing with partially unreliable dynamic random access memory-approximate DRAM

M Jung, DM Mathew, C Weis, N Wehn - Proceedings of the 53rd Annual …, 2016‏ - dl.acm.org
In the context of approximate computing, Approximate Dynamic Random Access Memory
(ADRAM) enables the tradeoff between energy efficiency, performance and reliability. The …

Omitting refresh: A case study for commodity and wide i/o drams

M Jung, É Zulian, DM Mathew, M Herrmann… - Proceedings of the …, 2015‏ - dl.acm.org
Dynamic Random Access Memories (DRAM) have a big impact on performance and
contribute significantly to the total power consumption in systems ranging from mobile …

Efficient reliability management in SoCs-an approximate DRAM perspective

M Jung, DM Mathew, C Weis… - 2016 21st Asia and South …, 2016‏ - ieeexplore.ieee.org
In today's computing systems Dynamic Random Access Memories (DRAMs) have a large
influence on performance and contribute significantly to the total power consumption. Thus …

Using run-time reverse-engineering to optimize DRAM refresh

DM Mathew, ÉF Zulian, M Jung, K Kraft… - Proceedings of the …, 2017‏ - dl.acm.org
The overhead of DRAM refresh is increasing with each density generation. To help offset
some of this overhead, JEDEC designed the modern Auto-Refresh command with a highly …

ADROIT: An Adaptive Dynamic Refresh Optimization Framework for DRAM Energy Saving In DNN Training

X Lin, L Sun, F Tu, L Liu, X Li, S Wei… - 2021 58th ACM/IEEE …, 2021‏ - ieeexplore.ieee.org
To achieve high accuracy, DNN training usually consumes and generates myriads of data,
which requires a large DRAM for efficient processing. The refresh power consumption in …

Cross-layer refresh mitigation for efficient and reliable DRAM systems: A comparative study

X Ding, X Liang, Y Li - 2017 IEEE International Test …, 2017‏ - ieeexplore.ieee.org
DRAM is a crucial component in computing systems, and is expected to be even more
important as data-intensive applications become more prominent. A key challenge in …

[PDF][PDF] Efficient fault tolerance for selected scientific computing algorithms on heterogeneous and approximate computer architectures

A Schöll - 2018‏ - core.ac.uk
Simulation technology and scientific computing play an essential role in the majority of
scientific domains and have become established techniques to solve central challenges in …

Intelligent Vision Systems: Exploring the State-of-the-Art and Opportunities for the Future

S Advani, S Kestur, V Narayanan - 2015 IEEE International …, 2015‏ - ieeexplore.ieee.org
Vision and video applications are becoming pervasive in mobile and embedded systems.
Consumer wearable devices require capabilities for real-time video analytics and prolonged …