Konstantinos Balaskas

Konstantinos Balaskas

Auto-generated profile, based on publicly available information

Konstantinos Balaskas

University of Patras

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 January 2024 in IEEE Transactions on Emerging Topics in Computing
Verified
Journal Article
Closed access
KBKonstantinos Balaskas
Konstantinos Balaskas
AKAndreas Karatzas
Andreas Karatzas
CSChristos Sad
Christos Sad
Kostas SioziosKostas Siozios
Kostas Siozios
IAIraklis Anagnostopoulos
Iraklis Anagnostopoulos
JHJ¨org Henkel
J¨org Henkel
Deep Neural Networks (DNNs) have shown significant advantages in a wide variety of domains. However, DNNs are becoming computationally intensive and energy hungry at an exponential pace, while at the same time, there is a vast demand for running sophisticated DNN-based services on resource constrained embedded devices. In this paper, we target energy-efficient inference on embedded DNN accelerators. To that end, we propose an automated framework to compress DNNs in a hardware-aware manner by jointly employing pruning and quantization. We explore, for the first time, per-layer fine- and coarse-grained pruning, in the same DNN architecture, in addition to low bit-width mixed-precision quantization for weights and activations. Reinforcement Learning (RL) is used to explore the associated design space and identify the pruning-quantization configuration so that the energy consumption is minimized whilst the prediction accuracy loss is retained at acceptable levels. Using our novel composite RL agent we are able to extract energy-efficient solutions without requiring retraining and/or fine tuning. Our extensive experimental evaluation over widely used DNNs and the CIFAR-10/100 and ImageNet datasets demonstrates that our framework achieves 39% average energy reduction for 1.7% average accuracy loss and outperforms significantly the state-of-the-art approaches.

1 September 2021 in IEEE Transactions on Circuits and Systems I: Regular Papers
Verified
Journal Article
Closed access
KBKonstantinos Balaskas
Konstantinos Balaskas
GZGeorgios Zervakis
Georgios Zervakis
HAHussam Amrouch
Hussam Amrouch
JHJorg Henkel
Jorg Henkel
Kostas SioziosKostas Siozios
Kostas Siozios
Transistor aging phenomena manifest themselves as degradations in the main electrical characteristics of transistors. Over time, they result in a significant increase of cell propagation delay, leading to errors due to timing violations, since the operating frequency becomes unsustainable as the circuit ages. Conventional techniques employ timing guardbands to mitigate aging-induced delay increase, which leads to considerable performance losses from the beginning of the circuit’s lifetime. Leveraging the inherent error resilience of a vast number of application domains, approximate computing was recently introduced as an aging mitigation mechanism. In this work, we present the first automated framework for generating aging-aware approximate circuits . Our framework, by applying directed gate-level netlist approximation, induces a small functional error and recovers the delay degradation due to aging. As a result, our optimized circuits eliminate aging-induced timing errors. Experimental evaluation over a variety of arithmetic circuits and image processing benchmarks demonstrates that for an average error of merely $5\times 10^{-3}$ , our framework completely eliminates aging-induced timing guardbands. Compared to the respective baseline circuits without timing guardbands (i.e., iso-performance evaluation), the error of the circuits generated by our framework is $1208\times $ smaller.

23 June 2022 in IEEE Transactions on Circuits and Systems I: Regular Papers
Verified
Journal Article
Closed access
KBKonstantinos Balaskas
Konstantinos Balaskas
FKFlorian Klemme
Florian Klemme
GZGeorgios Zervakis
Georgios Zervakis
Kostas SioziosKostas Siozios
Kostas Siozios
HAHussam Amrouch
Hussam Amrouch
JHJorg Henkel
Jorg Henkel
One of the major barriers that CMOS devices face at nanometer scale is increasing parameter variation due to manufacturing imperfections. Process variations severely inhibit the reliable operation of circuits, as the operational frequency at the nominal process corner is insufficient to suppress timing violations across the entire variability spectrum. To avoid variability-induced timing errors, previous efforts impose pessimistic and performance-degrading timing guardbands atop the operating frequency. In this work, we employ approximate computing principles and propose a circuit-agnostic automated framework for generating variability-aware approximate circuits that eliminate process-induced timing guardbands. Variability effects are accurately portrayed with the creation of variation-aware standard cell libraries, fully compatible with standard EDA tools. The underlying transistors are fully calibrated against industrial measurements from Intel 14nm FinFET in which both electrical characteristics of transistors and variability effects are accurately captured. In this work, we explore the design space of approximate variability-aware designs to automatically generate circuits of reduced variability and increased performance without the need for timing guardbands. Experimental results show that by introducing negligible functional error of merely $\boldsymbol {5.3 \times 10^{-3}}$ , our variability-aware approximate circuits can be reliably operated under process variations without sacrificing the application performance.

5 December 2024 in IEEE Embedded Systems Letters
Verified
Journal Article
Closed access
KBKonstantinos Balaskas
Konstantinos Balaskas
HKHeba Khdr
Heba Khdr
MSMohammed Bakr Sikal
Mohammed Bakr Sikal
FKFabian Kreß
Fabian Kreß
Kostas SioziosKostas Siozios
Kostas Siozios
JHJörg Henkel
Jörg Henkel
The significant advancements of deep neural networks (DNNs) in a wide range of application domains have spawned the need for more specialized, sophisticated solutions in the form of multi-DNN workloads. Heterogeneous DNN accelerators have emerged as an elegant solution to tackle the workloads’ inherent diversity, achieving significant improvements compared to homogeneous solutions. However, utilizing off-the-shelf architectures provides suboptimal adaptability to given workloads, whereas custom design approaches offer limited heterogeneity, and thus reduced gains. In this letter, we combat these shortcomings and propose an exploration-based framework to holistically design heterogeneous accelerators, tailored for multi-DNN workloads. Our framework is workload-agnostic and leverages architectural heterogeneity to its full potential, by integrating low-precision arithmetic and custom structural parameters. We explore the formed design space, targeting to minimize the system’s energy-delay product (EDP) via heuristic techniques. Our proposed accelerators achieve, on average, a significant $5.5\times $ reduction in EDP compared to the state of the art across various multi-DNN workloads.

29 November 2024 in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Verified
Journal Article
Closed access
KPKilian Pfeiffer
Kilian Pfeiffer
KBKonstantinos Balaskas
Konstantinos Balaskas
Kostas SioziosKostas Siozios
Kostas Siozios
JHJörg Henkel
Jörg Henkel
In Federated Learning (FL), devices that participate in the training usually have heterogeneous resources, i.e., energy availability. In current deployments of FL, devices that do not fulfill certain hardware requirements are often dropped from the collaborative training. However, dropping devices in FL can degrade training accuracy and introduce bias or unfairness. Several works have tackled this problem on an algorithm level, e.g., by letting constrained devices train a subset of the server neural network (NN) model. However, it has been observed that these techniques are not effective w.r.t. accuracy. Importantly, they make simplistic assumptions about devices’ resources via indirect metrics, such as multiply accumulate (MAC) operations or peak memory requirements. We observe that memory access costs (that are currently not considered in simplistic metrics) have a significant impact on the energy consumption. In this work, for the first time, we consider on-device accelerator design for FL with heterogeneous devices. We utilize compressed arithmetic formats and approximate computing, targeting to satisfy limited energy budgets. Using a hardware-aware energy model, we observe that, contrary to the state of the art’s moderate energy reduction, our technique allows for lowering the energy requirements (by $4\times $ ) while maintaining higher accuracy.

1 September 2025 in IEEE Transactions on Circuits and Systems for Artificial Intelligence
Verified
Journal Article
Closed access
VMVojtech Mrazek
Vojtech Mrazek
KBKonstantinos Balaskas
Konstantinos Balaskas
PDPaula Carolina Lozano Duarte
Paula Carolina Lozano Duarte
ZVZdenek Vasicek
Zdenek Vasicek
MTMehdi Tahoori
Mehdi Tahoori
GZGeorgios Zervakis
Georgios Zervakis
Printed electronics offer a promising alternative for applications beyond silicon-based systems, requiring properties like flexibility, stretchability, conformality, and ultra-low fabrication costs. Despite the large feature sizes in printed electronics, printed neural networks have attracted attention for meeting target application requirements, though realizing complex circuits remains challenging. This work bridges the gap between classification accuracy and area efficiency in printed neural networks, covering the entire processing-near-sensor system design and co-optimization from the analog-to-digital interface– a major area and power bottleneck–to the digital classifier. We propose an automated framework for designing printed Ternary Neural Networks with arbitrary input precision, utilizing multi-objective optimization and holistic approximation. Our circuits outperform existing approximate printed neural networks by 17× in area and 59× in power on average, being the first to enable printed-battery-powered operation with under 5% accuracy loss while accounting for analog-to-digital interfacing costs.