Chair News

Published on

The joint IEEE/ACM HPCA/CGO/PPoPP/CC 2026 conference took place in Sydney, Australia, from Saturday, 31 January to Wednesday, 4 February 2026, bringing together researchers from architecture, compilers, and parallel programming. Our Chair was represented by Prof. Anderson Faustino da Silva, Asif Ali Khan, and João Paulo C. de Lima, who contributed across several co-located venues. At CGO 2026, Anderson Faustino da Silva presented two papers in collaboration with Prof. Fernando Pereira: On the Precision of Dynamic Program Fingerprints Based on Performance Counters and Binary Diffing via Library Signatures. At HPCA 2026, João presented Count2Multiply: Reliable In-Memory High-Radix Counting, the result of a long-standing collaboration with Prof. Alex K. Jones. In addition, João gave an invited talk titled “Fault Injection Framework for Processing-using-Memory Architectures” at MCCSys 2026, the Workshop on Memory-Centric Computing Systems, and Asif presented “Compilation for Heterogeneous Near-Memory and In-Memory Computing Architectures” at LATHC 2026, the Workshop on Languages, Tools, and Compilers for Heterogeneous Computing.

 

Published on

The CC chair welcomes Tauseef Ahmed to the team! Tauseef obtained his Bachelor’s degree in Electronics and Communication from the PES University in Bangalore India in 2013. He then spent several years working for companies such as Siemens, Intel, and IMEC. In 2025 he finished his Master’s degree in Artificial Intelligence from the Maastricht University. Tauseef joins the team in the context of the REACT Doctoral Network financed by the European Union within the Marie Skłodowska-Curie Actions (MSCA). This project looks at self-awaRe nEuromorphic ArChiTectures, from the point of view of security, reliability and energy-efficiency. Tauseef experience in the micro-electronics industry and his background on computer architecture and artificial intelligence makes him ideal to tackle the challenges of programming and compiling reactive applications that execute close or at sensing nodes in cyber-physical systems (CPS). We look forward to working with Tauseef and are excited about having him in the team! 

Published on

The CC chair congratulates Robert Khasanov and Till Smejkal for receiving a Best Paper Award (Honorable Mention) at the 26th ACM/IFIP International Middleware Conference, held from December 15 to December 19, 2025, in Nashville, USA.

This award was presented for the paper “HARP: Energy-Aware and Adaptive Management of Heterogeneous Processors”, which is a result of a long-standing collaboration with Prof. Hermann Härtig and the Chair of Operating Systems. Robert Khasanov and Till Smejkal contributed equally to this work; the paper was presented at the conference by Robert Khasanov.

The paper introduces HARP, a resource-management framework for heterogeneous processors that enables coordinated adaptivity between the operating system and applications, leading to improved performance and energy efficiency.

The annual ACM/IFIP International Middleware Conference is a major international forum for research on middleware and distributed systems, spanning topics from distributed platforms to resource management and energy-aware computing.

TU Dresden news from Dec 19, 2025

Published on

TUD Dresden University of Technology performed exceptionally well in the 2024 call for Marie Skłodowska-Curie Actions (MSCA) Doctoral Networks. In total, seven projects involving TU Dresden (TUD) were selected for funding. One of these networks is coordinated by TUD. In the other six, TUD is involved as a partner. A total of around EUR 2.76 million will be allocated to TUD for this purpose.

cfaed is part of this success: TU Dresden holds a beneficiary partner role in the project 'REACT – Self-AwaRe NEuromorphic ArChiTectures: Security, Reliability and Energy-Efficiency'. The project lead on the TUD side lies in the hands of Prof. Jeronimo Castrillon (Faculty of Computer Science).

The REACT network (Funding sum for TUD: EUR 290,272.32) is training 15 doctoral students in neuromorphic systems—computer architectures inspired by the human brain. The goal is to develop adaptive, secure, and extremely energy-efficient hardware and software solutions. TU Dresden contributes its expertise in compiler technology and system software, helping to unlock the potential of neuromorphic systems for future, sustainable IT applications.

Published on

portrait photo of prof. Jeronimo Castrillon
Prof. Jeronimo Castrillon © M. Hahndorf for Collaborative Research Center/Transregio "Active-3D"

Jerónimo Castrillón-Mazo, Chair of Compiler Construction at the Faculty of Computer Science at TUD Dresden University of Technology (TUD), has been awarded one of the European Union's highly endowed ERC Consolidator Grants. His project, COMpilers for ExTreme Heterogeneity (COMETH), which will run for five years and receive EUR 2 million in funding, aims to develop a new generation of tools that will simplify the programming of increasingly complex computer systems for scientists, engineers, and other users.

Published on

The CC chair is happy to welcome Christian Schwarz to the team. Christian obtained his bachelor’s and master’s degrees from the TU Dresden in 2021 and 2025, respectively. He is an experience software developer in areas that span databases, operating systems and compilers. He did his master thesis with Prof. Horst Schirmeier at the department of Computer Science on “Scheduling Heterogeneous Memory for Database Query Execution”. Christian will join our activities on Machine Learning for Compilers, building on prior work from our group by Alexander Brauckmann and Anderson da Silva, and recent joint work with Fernando M Q Pereira. Concretely, he will be joining in the TransT DFG project on Transcending Tuning: Cross-domain Cross-platform Transfer Learning and Auto-tuning. We are excited about Christian’s future work and look forward to working with him!

Published on

The Chair for Compiler Construction hosted the first IEEE Cross-disciplinary Conference on Memory-Centric Computing (CCMCC), held on October 8-10, 2025, in Dresden. Our group was represented by Prof. Jeronimo Castrillon, Asif Ali Khan, Joao Paulo Cardoso de Lima, Anderson Faustino da Silva, Robert Khasanov, and Alexander Kusnezoff, with valuable organizational support from Christina Norkus and Conny Okuma. It was a pleasure to contribute to the event as part of the organizing team, serving as Local Organizing Chair and Finance Chair. During the CCMCC, Joao Paulo presented our paper titled "Efficient In-Memory Acceleration of Sparse Block Diagonal LLMs". The work introduces an automated framework with novel mapping and scheduling strategies to accelerate sparse LLM inference on compute-in-memory accelerators by exploiting block-diagonal sparsity.

Many thanks to Prof. Nima TaheriNejad, the General Chair, and to everyone who contributed to making this first edition a success. We are already looking forward to the next CCMCC, which will take place in Lyon, France, in October 2026.

Published on

This year’s ESWEEK was held in Taipei, Taiwan, and the Chair for Compiler Construction showed up in full force, with Prof. Jeronimo Castrillon, Asif Ali Khan, Joao Paulo Cardoso De Lima, and Hamid Farzaneh representing the group. We had quite an eventful week:

Sunday: We hosted a tutorial on “Hardware-Aware Compilation and Simulation for In-Memory Computing” together with X. Sharon Hu (University of Notre Dame) and Corey Lammie (IBM Zürich), as well as Asif, Hamid, and Joao Paulo from our team. Special thanks to Hadjer Benmeziane, William Simon, Yiyu Shi, Zheyu Yan, and Abu Sebastian for their valuable contribution to this tutorial. Learn more about the tools featured in the tutorial: Cinnamon and aihwkit.

Monday-Wednesday: Prof. Castrillon chaired the CASES international conference together with Christophe Dubach.

Monday: Shaokai Lin presented “Quasi-Static Scheduling for Deterministic Timed Concurrent Models on Multi-Core Hardware” at EMSOFT.

Wednesday: Prof. Castrillon participated in a panel discussion on “In-Memory Computing for Embedded Systems: a Hype or a Reality?” organized by Jian-Jia Chen.

Thursday: Prof. Castrillon organized a workshop on “Time-Centric Reactive Software (TCRS)” together with Hokeun Kim, where work by Hasna Bouraoui entitled "Combining Early Exit and Selective Prediction for Convolutional Neural Networks" was presented. On the same day, Joao Paulo gave a talk on “Cross-Layer Reliability Evaluation for Logic-in-Memory” at the International Workshop on Memory and Storage Computing (MSC).

We’re already looking forward to ESWEEK 2026 in Barcelona and preparing our submissions.

Published on

Asif Ali Khan and João Paulo de Lima represented the CC Chair at the 62nd Design Automation Conference (DAC'25), held from June 22nd to 25th in San Francisco, California. João gave a presentation and participated in a poster session, showcasing joint research with Prof. M. Hassan Najafi and Prof. Sercan Aygun on an error-tolerant computing method based on emerging non-volatile memories. Their paper, titled All-in-Memory Stochastic Computing using ReRAM, introduces a complete in-memory stochastic computing flow using ReRAM devices, handling random number generation, computation, and binary conversion all within memory. By exploiting ReRAM's physical properties and stochastic computing's error tolerance, it reduces energy and improves throughput compared to existing CMOS and ReRAM-based solutions. In addition, Asif, who was part of the technical program committee this year, was recognized with an Outstanding Reviewer Award.

Published on

We had a fantastic evening at this year’s Lange Nacht der Wissenschaften at TU Dresden! Many curious minds joined us to explore how intelligent compilers can make software faster and more efficient, and how novel computing approaches bring processing directly into memory to overcome traditional bottlenecks. From hands-on demos to deep tech discussions, it was exciting to share how compilers can bring different topics together — connecting high-level programming, optimization, and hardware innovation.