Integrating an open-source soft-GPU overlay with RISC-V control and high-bandwidth memory

Autor(en)

Muñoz Hernandez, Hector Gerardo, Taheri, Mahdi, Ali, Muhammad, Shahin, Keyvan, Syavashi, Alireza, Göhringer, Diana, Reichenbach, Marc, Herglotz, Christian, Hübner, Michael

Publikationsart

Scientific Article ref.

Erscheinungsjahr

2026

Verlag

Amsterdam : Elsevier BV

Freie Schlagworte

FPGA; Soft-core GPU; System-on-chip; RISC-V; HBM; Roofline

Quelle

Journal of systems architecture, S. 1 - 13

Band/Jahrgang

175

ISSN

1383-7621

DOI

https://doi.org/10.1016/j.sysarc.2026.103736

---

Reliability-aware performance optimization of DNN HW accelerators through heterogeneous quantization

Autor(en)

Nazari, Samira, Taheri, Mahdi, Azarpeyvand, Ali, Afsharchi, Mohsen, Herglotz, Christian, Jenihhin, Maksim

Publikationsart

Conference Proceeding refering

Erscheinungsjahr

2025

Verlag

New York : Institute of Electrical and Electronics Engineers (IEEE)

Quelle

2025 IEEE 26th Latin American Test Symposium (LATS), S. 1 - 6

ISBN

978-1-6654-7763-5

DOI

https://doi.org/10.1109/LATS65346.2025.10963948

---

FORTUNE : a negative memory overhead hardware-agnostic fault TOleRance techniqUe in DNNs

Autor(en)

Nazari, Samira, Taheri, Mahdi, Azarpeyvand, Ali, Afsharchi, Mohsen, Ghasempouri, Tara, Herglotz, Christian, Daneshtalab, Masoud, Jenihhin, Maksim

Publikationsart

Conference Proceeding refering

Erscheinungsjahr

2025

Verlag

New York : Institute of Electrical and Electronics Engineers (IEEE)

Quelle

2024 IEEE 33rd Asian Test Symposium (ATS), S. 1 - 6

ISBN

979-8-3315-2917-8

DOI

https://doi.org/10.1109/ATS64447.2024.10915463

---

SAFFIRA : a framework for assessing the reliability of systolic-array DNN accelerators

Autor(en)

Pappalardo, Salvatore, Bellarmino, Nicolo’, Deveautour, Bastien, Bosio, Alberto, Taheri, Mahdi, Daneshtalab, Masoud, Raik, Jaan, Jenihhin, Maksim

Publikationsart

Scientific Article ref.

Erscheinungsjahr

2025

Verlag

Singapore : World Scientific

Freie Schlagworte

Hardware accelerator; Systolic array; Deep neural networks; Fault simulation reliability; Resilience assessment; Reliability metrics

Quelle

Journal of circuits, systems, and computers : JCSC

Band/Jahrgang

34

Ausgabe/Heft

18

ISSN

1793-6454

DOI

https://doi.org/https://doi.org/10.1142/S0218126625430017

---

Reliability-aware hyperparameter optimization for ANN-to-SNN conversion

Autor(en)

Sharifian, Saeed, Taheri, Mahdi, Rashtchi, Vahid, Azarpeyvand, Ali, Herglotz, Christian, Jenihhin, Maksim

Publikationsart

Conference Proceeding refering

Erscheinungsjahr

2025

Verlag

Podgorica, Montenegro : WiPiEC Journal

Freie Schlagworte

Deep neural networks; Spiking neural networks; Reliability; Edge applications; Safety-critical applications

Quelle

WiPiEC journal : works in progress in embedded computing journal : special issue, S. 1 - 7

DOI

https://doi.org/https://doi.org/10.64552/wipiec.v11i1.85

---

SHIELD : PSO-based hardware Trojan detection for efficient and low-cost defense

Autor(en)

Hosseini, Mostafa, Azarpeyvand, Ali, Taheri, Mahdi, Ghasempouri, Tara, Jenihhin, Maksim

Publikationsart

Conference Proceeding refering

Erscheinungsjahr

2025

Verlag

Piscataway, NJ : IEEE

Freie Schlagworte

Hardware Trojan; Test generation; Particle swarm optimization; Depth-first search

Quelle

IEEE 31st International Symposium on On-Line Testing and Robust System Design (IOLTS), S. 1 - 4

ISBN

979-8-3315-3334-2

DOI

https://doi.org/10.1109/IOLTS65288.2025.11116879

---

RL-agent-based early-exit DNN architecture search framework

Autor(en)

Taheri, Mahdi, Patne, Parth, Cherezova, Natalia, Mahani, Ali, Herglotz, Christian, Jenihhin, Maksim

Publikationsart

Conference Proceeding refering

Erscheinungsjahr

2025

Verlag

Piscataway, NJ : IEEE

Freie Schlagworte

Deep neural networks; Dynamic DNNs; Early exit

Quelle

IEEE 28th International Symposium on Design and Diagnostics of Electronic Circuits and Systems (DDECS) : proceedings, S. 145 - 148

ISBN

979-8-3315-2801-0

DOI

https://doi.org/10.1109/DDECS63720.2025.11006795

---

GENIE : GENetIc algorithm-based REliability assessment methodology for deep neural networks

Autor(en)

Nazari, Samira, Taheri, Mahdi, Azarpeyvand, Ali, Afsharchi, Mohsen, Herglotz, Christian, Jenihhin, Maksim

Publikationsart

Conference Proceeding refering

Erscheinungsjahr

2025

Verlag

Piscataway, NJ : IEEE

Freie Schlagworte

Deep neural networks; Parallel processing; Memory overhead; Reliability; DNN accelerator

Quelle

2025 11th International Conference on Computing and Artificial Intelligence (ICCAI), S. 1 - 8

ISBN

979-8-3315-2492-0

DOI

https://doi.org/10.1109/ICCAI66501.2025.00049

---