Poster

Bolt: Bridging the Gap between Auto-tuners and Hardware-native Performance

Jiarong Xing · Leyuan Wang · Shang Zhang · Jack Chen · Ang Chen · Yibo Zhu

Keywords: [ hardware-efficient ml ] [ systems for ml ] [ ml compilers and runtime ] [ efficient inference and model serving ]

[ Abstract ]
[ Paper PDF
 
Oral presentation: Efficient Inference and Model Serving
Wed 31 Aug 2:15 p.m. PDT — 4:03 p.m. PDT

Abstract:

Today’s auto-tuners (e.g., AutoTVM, Ansor) generate efficient tensor programs by navigating a large search space to identify effective implementations, but they do so with opaque hardware details. Thus, their performance could fall behind that of hardware-native libraries (e.g., cuBLAS, cuDNN), which are hand-optimized by device vendors to extract high performance. On the other hand, these vendor libraries have a fixed set of supported functions and lack the customization and automation support afforded by auto-tuners. Bolt bridges this gap and achieves the best of both worlds by using hardware-native templated search, which is enabled by the recent trend that vendor libraries (e.g., CUTLASS) are increasingly modularized and reconfigurable. Bolt provides new opportunities to rethink end-to-end tensor optimizations at the graph, operator, and model levels. We demonstrate this concept by prototyping in TVM on NVIDIA GPUs—both in large deployment in our production environment. Our experiments show that Bolt can improve the inference speed of common convolutional neural networks by 2.5x on average over the state of the art, and it auto-tunes these models within 20 minutes.

Chat is not available.