Optimize Deep Learning GPU Operators with TVM: A Depthwise Convolution Example

Posted by crowwork on Tuesday August 22, 2017 @02:12PM
from the tvmlang.org dept.

http://www.tvmlang.org/2017/08/22/Optimize-Deep-Learning-GPU-Operators-with-TVM-A-Depthwise-Convolution-Example.html

( 3 comments | 24 points | on HN )

Re:Optimize Deep Learning GPU Operators with TVM: A Depthwis... by ysh329 on Wednesday August 23, 2017 @09:59PM (#15087034)

A collection about depthwise conv:

Optimize Deep Learning GPU Operators with TVM: A Depthwise Convolution Example http://tvmlang.org/2017/08/22/Optimize-Deep-Learning-GPU-Ope...

MobileNet and Depthwise Separable Convolution · Issue #70 · changtimwu/changtimwu.github.com https://github.com/changtimwu/changtimwu.github.com/issues/7...

yonghenglh6/DepthwiseConvolution: A personal mobile convolution layer implementation on caffe by liuhao.(only GPU) https://github.com/yonghenglh6/DepthwiseConvolution

[ Parent | on HN ]

Re:Optimize Deep Learning GPU Operators with TVM: A Depthwis... by LaPrometheus on Tuesday August 22, 2017 @02:39PM (#15075230)

This is super cool. Any comparison with Weld? https://weld-project.github.io/

[ Parent | on HN ]

Re:Optimize Deep Learning GPU Operators with TVM: A Depthwis... by crowwork on Tuesday August 22, 2017 @02:54PM (#15075349)

as far as i know weld is optimized for data analytics workload so far(dataframe) while tvm is optimized for tensor deep learning workloads with great cpu, gpu and other supports

[ Parent | on HN ]