Scholar
Kaustubh Shivdikar
Google Scholar ID: NCTXsGMAAAAJ
AMD - Hardware Architect
Citations & Impact
Citations
248
H-index
7
i10-index
5
Publications
20
Co-authors
32
list available
Contact
No contact links provided.
Publications
Resume (English only)
Academic Achievements
- Published 26 papers with 255 citations. Some of the research works include:
- - Characterizing the Behavior of Training Mamba-based State Space Models on GPUs
- - FIDESlib: A Fully-Fledged Open-Source FHE Library for Efficient CKKS on GPUs
- - PIMnet: A Domain-Specific Network for Efficient Collective Communication in Scalable PIM
- - Scalability Limitations of Processing-in-Memory using Real System Evaluations
- - MaxK-GNN: Extremely Fast GPU Kernel Design for Accelerating Graph Neural Networks Training
- - NeuraChip: Accelerating GNN Computations with a Hash-based Decoupled Spatial Accelerator
Research Experience
- Currently a Sr. Silicon Design Engineer at AMD, working on designing Graph Neural Networks.
Education
- Degree: Ph.D.; University: Not specified; Advisor: David Kaeli (Lab Name).
Background
- Research Interests: Graph Neural Networks, hardware accelerators, DNN Security, Sparse Matrices, GPU Computing, Computer Architecture. Professional Field: Designing a spatial accelerator for graph workloads.
Miscellany
- Currently working in the Hardware Research department at AMD as a Sr. Silicon Design Engineer.
