Invention Grant
- Patent Title: Dynamic operation allocation for neural networks
-
Application No.: US15721716Application Date: 2017-09-29
-
Publication No.: US10585703B2Publication Date: 2020-03-10
- Inventor: Francesco Rossi , Gaurav Kapoor , Michael R. Siracusa , William B. March
- Applicant: Apple Inc.
- Applicant Address: US CA Cupertino
- Assignee: APPLE INC.
- Current Assignee: APPLE INC.
- Current Assignee Address: US CA Cupertino
- Agency: Morgan, Lewis & Bockius LLP
- Main IPC: G06F8/41
- IPC: G06F8/41 ; G06F9/50 ; G06N3/063 ; G06N3/02

Abstract:
The subject technology provides for dynamic task allocation for neural network models. The subject technology determines an operation performed at a node of a neural network model. The subject technology assigns an annotation to indicate whether the operation is better performed on a CPU or a GPU based at least in part on hardware capabilities of a target platform. The subject technology determines whether the neural network model includes a second layer. The subject technology, in response to determining that the neural network model includes a second layer, for each node of the second layer of the neural network model, determines a second operation performed at the node. Further the subject technology assigns a second annotation to indicate whether the second operation is better performed on the CPU or the GPU based at least in part on the hardware capabilities of the target platform.
Public/Granted literature
- US20180349189A1 DYNAMIC TASK ALLOCATION FOR NEURAL NETWORKS Public/Granted day:2018-12-06
Information query