Bahar Asgari
Assistant Professor
Designs domain-specific accelerators and near-data processing architectures for sparse workloads such as graph analytics, recommendation and deep learning.
Bahar Asgari's lab at UMD works on accelerators, chips & architecture and recommendation, search & web. 50 papers since 2023, 9 at top venues, mostly MICRO and ASPLOS. 7 PhD students and 1 master's student. 12 alumni, most often to the Department of Computer Science. He is interested in different areas of computer systems and UMD. She completed her undergraduate studies in computer science at UMD in December of 2023 with minors in Business Analytics and Persian Studies. Hana’s research with CASL focused on accelerating recommendation systems.
At a glance
- papers, 2023–2026
- 13/ year
- at T1 venues
- 18%
- PhD students
- 7
- MS students
- 1
- alumni listed
- 12
Papers by yearfrom the lab homepage and DBLP, 2023–2026
- 14 papers2026 · T1 1 · collecting
- 21 papers2025 · T1 6
- 6 papers2024 · T1 2
- 9 papers2023 · T1 0
T1 venuesOtherVenues: MICRO, ASPLOS, DAC, NeurIPS
Research focuspaper titles that use each topic's words
- Accelerators, chips & architecture9 papers
- Recommendation, search & web0 papers
- graph analytics0 papers
- domain-specific accelerators0 papers
- workloads0 papers
Paper list
From the lab homepage and DBLP.
202614 papers
- —Arcalís: Accelerating Remote Procedure Calls Using a Líghtweight Near-Cache Solution (MICRO)
- —Diphda: Demystifying Processing-in-HBM for SpGEMM across Diverse Data Mapping (IISWC)
- —Márohu: Scalable CGRA Mapping through Reinforcement Learning a Hierarchical Pointer Network (Y'arch at ISCA)
- —Orion: Online Reduction of Activation Dispatch Overhead in Disaggregated MoE Inference (MLArchSys at ISCA)
- —Polaris: Enabling PIM-Optimized Low-Energy SpGEMM using a Remote-Access-Aware Mapping Solution (DAC LBR)
- —Procyon: Promoting Fine-Grained Multi-Tenancy to Optimize Sparse Streaming Accelerators (DAC)
- IEEE Comput. Archit. Lett.Sāmaya: Safe Memory Accesses to Expand Coarse-Grain Reconfigurable Array Adoption (opens in a new tab)
- —Sāmaya: Safe Memory Accesses to Expand Coarse-Grain Reconfigurable Array Adoption (IEEE CAL)
- ISPASSSitula: Studying the Interplay of Sparse Formats and CPU/GPU Libraries (opens in a new tab)
- —Situla: Studying the Interplay of Sparse Formats and CPU/GPU Libraries (ISPASS)
- MICRO (T1)Special Issue on AI for Hardware and Hardware for AI: A Coevolution Shaping the Future of Computing (opens in a new tab)
- —Towards Transparent and Reproducible Accelerator Research (OSCAR at ISCA)
202521 papers
- IISWCBelenos: Bottleneck Evaluation to Link Biomechanics to Novel Computing Optimizations (opens in a new tab)
- —Belenos: Bottleneck Evaluation to Link Biomechanics to Novel Computing Optimizations (IISWC)
- MICRO (T1)Bootes: Boosting the Efficiency of Sparse Accelerators Using Spectral Clustering (opens in a new tab)
- —Boötes: Boosting the Efficiency of Sparse Accelerators Using Spectral Clustering (MICRO)
- MICRO (T1)Chasoň: Supporting Cross HBM Channel Data Migration to Enable Efficient Sparse Algebraic Acceleration (opens in a new tab)
- —Chasoň: Supporting Cross-HBM Channel Data Migration to Enable Efficient Sparse Algebraic Acceleration (MICRO)
- MICRO (T1)Coruscant: Co-Designing GPU Kernel and Sparse Tensor Core to Advocate Unstructured Sparsity in Efficient LLM Inference (opens in a new tab)
- —Coruscant: Co-Designing GPU Kernel and Sparse Tensor Core to Advocate Unstructured Sparsity in Efficient LLM Inference (MICRO)
- IEEE Comput. Archit. Lett.DynaFlow: An ML Framework for Dynamic Dataflow Selection in SpGEMM Accelerators (opens in a new tab)
- —DynaFlow: An ML Framework for Dynamic Dataflow Selection in SpGEMM accelerators (IEEE CAL)
- IEEE Comput. Archit. Lett.Electra: Eliminating the Ineffectual Computations on Bitmap Compressed Matrices (opens in a new tab)
- ISPASSLa Superba: Leveraging a Self-Comparison Method to Understand the Performance Benefits of Sparse Acceleration Optimizations (opens in a new tab)
20246 papers
- MICRO (T1)Acamar: A Dynamically Reconfigurable Scientific Computing Accelerator for Robust Convergence and Minimal Resource Underutilization (opens in a new tab)
- —Acamar: A Dynamically Reconfigurable Scientific Computing Accelerator for Robust Convergence and Minimal Resource Underutilization (MICRO)
- —Electra: Eliminating the Ineffectual Computations of Bitmap Compressed Matrices (IEEE CAL)
- ASPLOS (T1)GUST: Graph Edge-Coloring Utilization for Accelerating Sparse Matrix Vector Multiplication (opens in a new tab)
- —GUST: Graph Edge-Coloring Utilization for Accelerating Sparse Matrix Vector Multiplication (ASPLOS)
- —Misam: Using ML in Dataflow Selection of Sparse-Sparse Matrix Multiplication (MLArchSys at ISCA)
20239 papers
- EDGEContext-Aware Task Handling in Resource-Constrained Robots with Virtualization (opens in a new tab)
- —Context-Aware Task Handling in Resource-Constrained Robots with Virtualization (IEEE EDGE)
- EDGECreating Robust Deep Neural Networks with Coded Distributed Computing for IoT (opens in a new tab)
- —Creating Robust Deep Neural Networks With Coded Distributed Computing for IoT (IEEE EDGE)
- —LCP: A Low-Communication Parallelization Method for Fast Neural Network Inference for IoT (CSCE-ICOMP)
- VLSI-SoCMemory-Based Computing for Energy-Efficient AI: Grand Challenges (opens in a new tab)
- —Memory-based computing for energy-efficient AI: Grand challenges (VLSI-SoC)
- EDGESpica: Exploring FPGA Optimizations to Enable an Efficient SpMV Implementation for Computations at Edge (opens in a new tab)
- —Spica: Exploring FPGA Optimizations to Enable an Efficient SpMV Implementation for Computations at Edge (IEEE EDGE)
Where alumni went12 listed, 60% to industry
Industry 3Academia 2
Most often to
- the Department of Computer Science. He is interested in different areas of computer systems
- UMD. She completed her undergraduate studies in computer science at UMD in December of 2023 with minors in Business Analytics and Persian Studies. Hana’s research with CASL focused on accelerating recommendation systems
Also to
- University of Maryland
Apply to UMDFall 2027
Not announced yet
UMD Computer Science PhD
- Round
- Single round
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- USNYU
Brandon Reagen
- Accelerators, chips & architecture
- CAToronto
Maryam Mehri Dehnavi
- Compilers & programming languages
- GBImperial
Christos Bouganis
- Accelerators, chips & architecture
- USStanford
Kunle Olukotun
- Accelerators, chips & architecture
- GBUCL
Adnan Mehonic
- Accelerators, chips & architecture
- CNSJTU
Fangxin Liu
- Accelerators, chips & architecture
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage and DBLP
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.