Sort by
Refine Your Search
-
and tuning. Moderate research project experience training large-scale foundation models, especially pipeline/model parallelism. Track record of creating HPC software for numerical methods. Domain
-
and tuning. Moderate research project experience training large-scale foundation models, especially pipeline/model parallelism. Track record of creating HPC software for numerical methods. Domain
-
-scale foundation models, especially pipeline/model parallelism. Track record of creating HPC software for numerical methods. Domain expertise in areas like computational fluid dynamics, material
-
them on massively parallel computers ? CPU, GPU, and APU systems. The successful candidate will work with Prof. Uri Shumlak and Prof. Jingwei Hu and contribute to a project that combines novel low-rank
-
with implementation machine learning methods and development of high-performance scientific software with CPU and GPU parallelization, knowledge of LAMMPS. 2. Computational investigations of catalytic