Thermal-aware design for approximate DNN accelerators
G Zervakis, I Anagnostopoulos… - IEEE Transactions …, 2022 - ieeexplore.ieee.org
Recent breakthroughs in Neural Networks (NNs) have made DNN accelerators ubiquitous
and led to an ever-increasing quest on adopting them from Cloud to edge computing …
and led to an ever-increasing quest on adopting them from Cloud to edge computing …
ALWANN: Automatic layer-wise approximation of deep neural network accelerators without retraining
The state-of-the-art approaches employ approximate computing to reduce the energy
consumption of DNN hardware. Approximate DNNs then require extensive retraining …
consumption of DNN hardware. Approximate DNNs then require extensive retraining …
Control variate approximation for DNN accelerators
G Zervakis, O Spantidi… - 2021 58th acm/ieee …, 2021 - ieeexplore.ieee.org
In this work, we introduce a control variate approximation technique for low error
approximate Deep Neural Network (DNN) accelerators. The control variate technique is …
approximate Deep Neural Network (DNN) accelerators. The control variate technique is …
Hardware approximate techniques for deep neural network accelerators: A survey
Deep Neural Networks (DNNs) are very popular because of their high performance in
various cognitive tasks in Machine Learning (ML). Recent advancements in DNNs have …
various cognitive tasks in Machine Learning (ML). Recent advancements in DNNs have …
LAMBDA: An open framework for deep neural network accelerators simulation
Many tasks in the realm of recognition, mining, and synthesis are increasingly being
implemented by using machine learning approaches. In particular, deep neural networks …
implemented by using machine learning approaches. In particular, deep neural networks …
A uniform modeling methodology for benchmarking dnn accelerators
Deep Neural Networks (DNNs) have achieved tremendous success in many application
domains. Inspired by its success, specialized accelerators have been and continue to be …
domains. Inspired by its success, specialized accelerators have been and continue to be …
CANN: Curable approximations for high-performance deep neural network accelerators
Approximate Computing (AC) has emerged as a means for improving the performance, area
and power-/energy-efficiency of a digital design at the cost of output quality degradation …
and power-/energy-efficiency of a digital design at the cost of output quality degradation …
Energy-efficient dnn inference on approximate accelerators through formal property exploration
O Spantidi, G Zervakis… - IEEE Transactions on …, 2022 - ieeexplore.ieee.org
Deep neural networks (DNNs) are being heavily utilized in modern applications, putting
energy-constraint devices to the test. To bypass high energy consumption issues …
energy-constraint devices to the test. To bypass high energy consumption issues …
Adapt: Fast emulation of approximate dnn accelerators in pytorch
Current state-of-the-art employs approximate multipliers to address the highly increased
power demands of deep neural network (DNN) accelerators. However, evaluating the …
power demands of deep neural network (DNN) accelerators. However, evaluating the …
Positive/negative approximate multipliers for DNN accelerators
O Spantidi, G Zervakis… - 2021 IEEE/ACM …, 2021 - ieeexplore.ieee.org
Recent Deep Neural Networks (DNNs) manage to deliver superhuman accuracy levels on
many AI tasks. DNN accelerators are becoming integral components of modern systems-on …
many AI tasks. DNN accelerators are becoming integral components of modern systems-on …