Jump to content

18412698. ARTIFICIAL INTELLIGENCE MODEL INFERENCE AND TUNING PEER-TO-PEER NETWORK SYSTEM (NVIDIA CORPORATION)

From WikiPatents


ARTIFICIAL INTELLIGENCE MODEL INFERENCE AND TUNING PEER-TO-PEER NETWORK SYSTEM

Organization Name

NVIDIA CORPORATION

Inventor(s)

Bing Xu of Kirkland WA US

ARTIFICIAL INTELLIGENCE MODEL INFERENCE AND TUNING PEER-TO-PEER NETWORK SYSTEM

This abstract first appeared for US patent application 18412698 titled 'ARTIFICIAL INTELLIGENCE MODEL INFERENCE AND TUNING PEER-TO-PEER NETWORK SYSTEM

Original Abstract Submitted

A peer-to-peer signal node may include a communication interface in communication with a plurality of client machines via a network. The communication interface may receive a client request message identifying a designated computing model and input data on which to execute the designated computing model. The peer-to-peer signal node may also include a computation node registry maintaining access information for a plurality of distributed computation nodes accessible via the network. The peer-to-peer signal node may also include a work scheduler, which may select a distributed computation node and transmit a computation request message to the selected node identifying the designated computing model and the input data and establishing a communication session between the distributed computation node and the client machine through which the distributed computation node transmits a result obtained based on executing the computing model using the input data.

Cookies help us deliver our services. By using our services, you agree to our use of cookies.