5 papers
Symbolic Basic Block Profiling for Machine Learning Kernels
Jingyu Qiu, Rongcui Dong, Sreepathi Pai
Current basic block profiling techniques obtain the count of executions of each basic block in a program using dynamic instrumentation. These profiling counters create runtime over…
Formalizing Linear Motion G-code for Invariant Checking and Differential Testing of Fabrication Tools
Yumeng He, Chandrakana Nandi, Sreepathi Pai
The computational fabrication pipeline for 3D printing is much like a compiler - users design models in Computer Aided Design (CAD) tools that are lowered to polygon meshes to be u…
A Few Fit Most: Improving Performance Portability of SGEMM on GPUs using Multi-Versioning
Robert Hochgraf, Sreepathi Pai
Hand-optimizing linear algebra kernels for different GPU devices and applications is complex and labor-intensive. Instead, many developers use automatic performance tuning (autotun…
Modeling Utilization to Identify Shared-Memory Atomic Bottlenecks
Rongcui Dong, Sreepathi Pai
Performance analysis is critical for GPU programs with data-dependent behavior, but models like Roofline are not very useful for them and interpreting raw performance counters is t…
An SMT Formalization of Mixed-Precision Matrix Multiplication: Modeling Three Generations of Tensor Cores
Benjamin Valpey, Xinyi Li, Sreepathi Pai +1
Many recent computational accelerators provide non-standard (e.g., reduced precision) arithmetic operations to enhance performance for floating-point matrix multiplication. Unfortu…