Zhen Zheng
0009-0006-2692-713X
20 papers found
Refreshing results…
WiseGraph: Optimizing GNN with Joint Workload Partition of Graph and Operations
BladeDISC: Optimizing Dynamic Shape Machine Learning Workloads via Compiler Approach
Expanding the Edge: Enabling Efficient Winograd CNN Inference With Deep Reuse on Edge Device
RECom: A Compiler Approach to Accelerating Recommendation Model Inference with Massive Embedding Columns
Auto-Parallelizing Large Models with Rhino: A Systematic Approach on Production AI Platform
Optimizing DNN Compilation for Distributed Training With Joint OP and Tensor Fusion
DREW: Efficient Winograd CNN Inference with Deep Reuse
AStitch: enabling a new multi-dimensional optimization space for memory-intensive ML training and inference on modern SIMT architectures
Whale: Efficient giant model training over heterogeneous ${$GPUs$}$
Understanding and bridging the gaps in current GNN performance optimizations
Exploring deep reuse in winograd CNN inference
DAPPLE: A pipelined data parallel approach for training large models
DISC: A dynamic shape compiler for machine learning workloads
Optimizing distributed training deployment in heterogeneous GPU clusters
Gopipe: a granularity-oblivious programming framework for pipelined stencil executions on gpu
Whale: Scaling deep learning model training to the trillions
Fusionstitching: boosting memory intensive computations for deep learning workloads
HiWayLib: A software framework for enabling high performance communications for heterogeneous pipeline computations
Versapipe: a versatile programming framework for pipelined computing on GPU
基于 Cupti 接口的典型 Gpu 程序负载特征分析
Missing publications? Search for publications with a matching author name.