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 ]

Aggregated results from TensorFlow crowd-benchmarking (time, accuracy, energy, cost, ...)

[ Collaborative unification of AI ], [ CK-Caffe2 / CK-Caffe ], [ CK-Tensorflow ], [ Android app ], [ CK intro, vision and crowd-tuning; YouTube lecture ]

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

All raw files Type DNN engine Network Batch size Num batches FWBW time (sec.) FW time (sec.) Chars Platform CPU GPGPU OS Fail? User Replay
1) 5d400eb5344edf41 cuda TensorFlow library (prebuilt, cuda)
(1.1.0)		 tf-benchmark-overfeat 5 5 0.005 

(0.005 ± 0.000)
3 repetitions 		0.002 

(0.002 ± 0.000)
3 repetitions 		Hewlett-Packard (HP Z640 Workstation) Intel(R) Xeon(R) CPU E5-2650 v3 @ 2.30GHz GeForce GTX 1080 Ubuntu 16.04.2 LTS No 5f780539ea4a803d
2) c800644bea8ab968 cuda TensorFlow library (prebuilt, cuda)
(1.1.0)		 tf-benchmark-alexnet 5 5 0.007 

(0.008 ± 0.000)
3 repetitions 		0.002 

(0.002 ± 0.000)
3 repetitions 		Hewlett-Packard (HP Z640 Workstation) Intel(R) Xeon(R) CPU E5-2650 v3 @ 2.30GHz GeForce GTX 1080 Ubuntu 16.04.2 LTS No 5f780539ea4a803d
3) eae936e5b33c3ca2 cuda TensorFlow library (prebuilt, cuda)
(1.1.0)		 tf-benchmark-googlenet 5 5 0.019 

(0.019 ± 0.000)
3 repetitions 		0.006 

(0.006 ± 0.000)
3 repetitions 		Hewlett-Packard (HP Z640 Workstation) Intel(R) Xeon(R) CPU E5-2650 v3 @ 2.30GHz GeForce GTX 1080 Ubuntu 16.04.2 LTS No 5f780539ea4a803d
4) aaba71e4c1a8c48a cpu TensorFlow library (prebuilt, cpu)
(1.8.0)		 tf-benchmark-alexnet 5 5 0.062 
6 repetitions 		0.019 
6 repetitions 		Micro-Star International Co., Ltd. 1.0 (MS-7B48) Intel(R) Core(TM) i7-8700K CPU @ 3.70GHz Ubuntu 18.04 LTS No 607fb4454e59b982
5) 599a9ac896ddbe43 cpu
(1.11.0)		 tf-benchmark-alexnet 5 5 0.078 

(0.080 ± 0.003)
3 repetitions 		0.023 

(0.023 ± 0.001)
3 repetitions 		Xen 4.1.2_115-908.790. (HVM domU) Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40GHz Ubuntu 16.04.2 LTS No -
6) 1ee99a32aa8b055c cpu TensorFlow library (prebuilt, cpu)
(1.1.0)		 tf-benchmark-alexnet 5 5 0.114 

(0.117 ± 0.003)
6 repetitions 		0.037 

(0.037 ± 0.000)
6 repetitions 		Hewlett-Packard (HP Z640 Workstation) Intel(R) Xeon(R) CPU E5-2650 v3 @ 2.30GHz Ubuntu 16.04.2 LTS No 5f780539ea4a803d
7) 8b60916e3594288a cpu TensorFlow library (prebuilt, cpu)
(1.1.0)		 tf-benchmark-alexnet 5 5 0.263 

(0.278 ± 0.022)
3 repetitions 		0.084 

(0.087 ± 0.004)
3 repetitions 		Microsoft Corporation 7.0 (Virtual Machine) Intel(R) Xeon(R) CPU E5-2673 v3 @ 2.40GHz Ubuntu 16.04.2 LTS No azure-dnn
8) 8222b3648f59d5bb cpu TensorFlow library (prebuilt, cpu)
(1.1.0)		 tf-benchmark-alexnet 5 5 0.889 

(1.000 ± 0.102)
9 repetitions 		0.269 

(0.323 ± 0.057)
9 repetitions 		LENOVO ThinkPad X240 (20ALCTO1WW) Intel(R) Core(TM) i5-4210U CPU @ 1.70GHz Windows 10 No fgg-lenovo2
9) 48659d52220b6174 cpu TensorFlow library (prebuilt, cpu)
(1.8.0)		 tf-benchmark-alexnet 5 100) By O.E.M. (To Be Filled By O.E.M.) Intel(R) Core(TM) i5-6600K CPU @ 3.50GHz Windows 10 Yes {your


Developed by Grigori Fursin           
Implemented as a CK workflow 		                         
   
                      Hosted at