Collective Knowledge Aggregator proof-of-concept
Crowd results Raw CK browser Graphs Reports Datasets Models Home
Select CK-powered unified experimental workflow:
[ Participated users, platforms, OS, CPU, GPU, GPGPU, NN, NPU ] [ How to participate ] [ Motivation (PPT) (PDF) ] [ Papers 1 , 2 , 3] [ Android app ] [ Collective training set ] [ Unified AI ]
[ Collaborative unification of AI ], [ CK-Caffe2 / CK-Caffe ], [ CK-Tensorflow ], [ Android app ], [ CK intro, vision and crowd-tuning; YouTube lecture ]

Type: DNN engine: Model:
Platform: CPU:
OS: GPGPU:

# Platform OS CPU GPGPU Type DNN engine Model Choices (env) FWBW
min time

(exp time)
stat. repetitions		 FW BW Per layer Model size Model accuracy on ImageNet Model topology and parameters Power consumption (W)
min / max		 Memory usage (MB) Bug detected? User Replay
1 LENOVO ThinkPad X240 (20ALCTO1WW) Windows 10 Intel(R) Core(TM) i5-4210U CPU @ 1.70GHz Intel(R) HD Graphics 4400 BVLC Caffe framework (opencl,libdnn,viennacl)

Version master-880111e

 resnet, resnet152 Batch size=10

 236 MB Yes fgg-lenovo2

[ See raw files ]
2 LENOVO ThinkPad X240 (20ALCTO1WW) Windows 10 Intel(R) Core(TM) i5-4210U CPU @ 1.70GHz BVLC Caffe framework (cpu)

Version trunk-90e8abc

 deepscale, squeezenet, 1.1 Batch size=1

 93 ms.

(93 ± 0 ms.)
3 repetitions 		55 ± 8 ms. 39 ± 8 ms. Stats per layer



 5 MB 0.58388 / 0.810123 32 MB fgg-lenovo2

[ See raw files ]
3 LENOVO ThinkPad X240 (20ALCTO1WW) Windows 10 Intel(R) Core(TM) i5-4210U CPU @ 1.70GHz BVLC Caffe framework (cpu)

Version trunk-90e8abc

 bvlc, googlenet Batch size=10

 2734 ms.

(2830 ± 124 ms.)
3 repetitions 		1781 ± 124 ms. 1070 ± 8 ms. Stats per layer



 54 MB 0.689299 / 0.891441 551 MB fgg-lenovo2

[ See raw files ]
4 LENOVO ThinkPad X240 (20ALCTO1WW) Windows 10 Intel(R) Core(TM) i5-4210U CPU @ 1.70GHz BVLC Caffe framework (cpu)

Version trunk-90e8abc

 resnet, resnet101 Batch size=1

 1093 ms.

(1134 ± 55 ms.)
3 repetitions 		586 ± 39 ms. 547 ± 16 ms. Stats per layer



 175 MB 354 MB fgg-lenovo2

[ See raw files ]
5 SAMSUNG SM-G930F Android 6.0.1 unknown-arm64-v8a BVLC Caffe framework (cpu)

Version trunk-c61d487

 bvlc, googlenet Batch size=1

 2866 ms.

(3037 ± 218 ms.)
3 repetitions 		1565 ± 135 ms. 1461 ± 134 ms. Stats per layer



 54 MB 0.689299 / 0.891441 deploy.prototxt 55 MB fgg-lenovo2

[ See raw files ]
6 LENOVO ThinkPad X240 (20ALCTO1WW) Windows 10 Intel(R) Core(TM) i5-4210U CPU @ 1.70GHz BVLC Caffe framework (cpu)

Version trunk-90e8abc

 bvlc, googlenet Batch size=1

 296 ms.

(306 ± 8 ms.)
3 repetitions 		172 ± 0 ms. 133 ± 8 ms. Stats per layer



 54 MB 0.689299 / 0.891441 55 MB fgg-lenovo2

[ See raw files ]
7 IBM SyNAPSE Ubuntu 14.04.4 LTS Intel(R) Xeon(R) CPU E5-2650 v2 @ 2.60GHz BVLC Caffe framework (cpu)

Version trunk-365ac88

 deepscale, squeezenet, 1.1 Batch size=8

 713 ms.

(713 ± 0 ms.)
3 repetitions 		323 ± 0 ms. 389 ± 1 ms. Stats per layer



 258 MB xuechao@cn.ibm.com

[ See raw files ]


Developed by Grigori Fursin           
Implemented as a CK workflow 		                         
   
                      Hosted at