1 paper
Frank Sifei Luan, Ron Yifeng Wang, Yile Gu +12
While ML model training and inference are both GPU-intensive, CPU-based data processing is often the bottleneck. Existing distributed data processing systems assume homogeneous res…