Josie E. Rodriguez Condia
Papers
8
Total Citations
67
H-Index
4
About
Josie E. Rodriguez Condia is a researcher specializing in the reliability and fault tolerance of Graphics Processing Units (GPUs) and Deep Neural Networks (DNNs), with a particular focus on safety-critical applications in automotive, robotics, and IoT domains. Her work sits at the intersection of computer architecture, hardware dependability, and artificial intelligence, addressing one of the most pressing challenges in modern computing: ensuring that AI-accelerated hardware remains trustworthy over extended operational lifetimes. Rodriguez Condia's most impactful contribution, "A Multi-level Approach to Evaluate the Impact of GPU Permanent Faults on CNN's Reliability" (2022, 32 citations), pioneered systematic methodologies for assessing how hardware degradation affects Convolutional Neural Network performance. She has extended this work to examine specialized accelerators such as Tensor Core Units, scheduling policies, and in-field GPU testing compaction methods. More recently, she has turned her attention to emerging paradigms like Split Computing, evaluating how DNNs partitioned across mobile and cloud environments maintain resilience under hardware faults. With publications consistently appearing since 2022 and accumulating over 60 citations, Rodriguez Condia represents a rising voice in dependable AI hardware research, producing work that directly informs the certification and deployment of intelligent systems in high-stakes environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Evaluating the Reliability of Supervised Compression for Split Computing8 citations · 2024
- 4A Compaction Method for STLs for GPU in-field test8 citations · 2022
- 5
- 6
- 7
- 8Enhancing the Reliability of Split Computing Deep Neural Networks2 citations · 2024