University of Arizona Reconfigurable Computing Lab (UA-RCL)
We research ways to bridge the gap between application engineers, hardware architects, and system software developers towards making emerging heterogeneous computing and neuromorphic computing systems accessible. For this, we are actively working on two thrusts: a) building intelligent runtime systems for hardware-agnostic application development and deployment on SoC scale heterogeneous computing systems; b) designing reconfigurable architectures for neuromorphic computing to perform both training and high-throughput inference on resource constraint environments.
News
- Our paper titled K-PACT: Kernel Planning for Adaptive Context Switching — A Framework for Clustering, Placement, and Prefetching in Spectrum Sensing is published in International Conference on Computer-Aided Design (ICCAD’25). The paper is available on (IEEE Xplore) and the source code is hosted on Github.
- Our paper titled A Portable Framework with Generalized Runtime Features for Task Graph Execution and Concurrent Multi-Application Deployment on Heterogeneous Systems (Science Direct) is published in Future Generation Computer Systems as journal preproof.
- Our paper titled HOPPERFISH: Holistic Profiling with Portable Extensible and Robust Framework Intended for Systems with Heterogeneity is accepted for publication in ACM Transactions on Architecture and Code Optimization and is available as just accepted in ACM. The source code is hosted on Github.
- Our paper titled Event-based Human Action Recognition: Rehabilitation Monitoring Dataset and Models is accepted for publication in IEEE International Conference on Future Machine Learning and Data Science (FMLDS’25).
- Our paper titled An Overview of Challenges and Requirements for Real-Time Spectrum Sensing in Modern RF Autonomy Systems (IEEE Xplore) is published in IEEE Design & Test.
- Our paper titled RIMMS: Runtime Integrated Memory Management System for Heterogeneous Computing is accepted for publication in Embedded Systems Week (ESWEEK’25) and is available as just accepted in ACM Transactions on Embedded Computing Systems as part of ESWEEK’25 special issue.
- Our paper titled A Unified Portable and Programmable Framework for Task-Based Execution and Dynamic Resource Management on Heterogeneous Systems (ACM) was published in Extreme Heterogeneity Solutions (ExHET’25).
- Our paper titled A Unified Portable and Programmable Framework for Task-Based Execution and Dynamic Resource Management on Heterogeneous Systems has won the Best Paper Award in Workshop on Extreme Heterogeneity Solutions (ExHET’25)! Congratulations Serhan! (Check out our blog post)
- Serhan Gener and Sahil Hassan held a Tutorial Session titled CEDR: A Holistic Software and Hardware Design Environment for Hardware Agnostic Application Development and Deployment on FPGA-Integrated Heterogeneous Systems in ISFPGA’25. (Check out our blog post).
- Our paper titled A Runtime Manager Integrated Emulation Environment for Heterogeneous SoC Design with RISC-V Cores (IEEE Xplore) was published in Heterogeneity in Computing Workshop (HCW’24).
Click here to check out archived news.
Open-Source Highlights
CEDR: a Compiler-integrated Extensible DSSoC Runtime
CEDR is an open-source, compiler-integrated, extensible DSSoC runtime that unifies a front-end compilation flow with a Linux-based runtime and intelligent scheduling framework. Applications are transformed into hardware-agnostic function calls, allowing the runtime to dynamically map tasks to the most suitable processing elements, enabling efficient resource utilization and productive deployment in heterogeneous systems.
RANC: Reconfigurable Architecture for Neuromorphic Computing
RANC is a highly flexible environment that enables rapid experimentation with neuromorphic architectures in both software via C++ simulation and hardware via FPGA emulation. RANC enables hardware architects and application engineers to investigate and tune parameters of their neuromorphic architecture that would otherwise be unavailable on a purely prefabricated ASIC.
K-PACT: Kernel Planning for Adaptive Context Switching
K-PACT is a software planning framework that clusters temporally independent kernels and places them onto reconfigurable hardware equipped with multiple IMEM banks, enabling nanosecond-scale context switching without requiring full hardware reconfiguration. This capability is particularly critical for dynamic workloads, where rapid adaptation and efficient reuse of resources are essential.
Highlights
Our Projects
Click the link below for our active and archived projects.
Our Team
Click the link below to meet our team members.