list.rst 3.89 KB
Newer Older
1
.. frameworks/validated/list.rst: 
2

3
.. _validated:
4 5


6
Validated Workloads
7
###################
8

9
We have validated performance [#f1]_ for the following workloads:
10

11 12 13 14 15 16 17
.. contents::
   :local:

.. _cpu_tensorflow:

CPU Tensorflow
==============
18

19
.. csv-table::
20
   :header: "TensorFlow Workload", "Genre of Deep learning"
21 22 23
   :widths: 27, 53
   :escape: ~

24 25 26 27
   Resnet50 v1, Image recognition
   Resnet50 v2, Image recognition
   Inception V3, Image recognition
   Inception V4, Image recognition
28 29
   Inception-ResNetv2, Image recognition
   MobileNet v1, Image recognition
30
   Faster RCNN, Object detection
31
   VGG16, Image recognition
32 33
   SSD-VGG16, Object detection
   SSD-MobileNetv1, Object detection
34
   R-FCN, Object detection
35
   Yolo v2, Object detection
36
   Transformer-LT, Language translation
37 38 39 40 41 42
   Wide & Deep, Recommender system
   NCF, Recommender system
   U-Net, Image segmentation
   DCGAN, Generative adversarial network
   DRAW, Image generation
   A3C, Reinforcement learning
43

44

45
.. _cpu_onnx:
46

47 48
CPU ONNX
========
49

L.S. Cook's avatar
L.S. Cook committed
50
Additionally, we validated the following workloads are functional through 
51
`nGraph ONNX importer`_. ONNX models can be downloaded from the `ONNX Model Zoo`_.
52 53

.. csv-table::
L.S. Cook's avatar
L.S. Cook committed
54
   :header: "ONNX Workload", "Genre of Deep Learning"
55 56 57
   :widths: 27, 53
   :escape: ~

58 59 60
   DenseNet-121, Image recognition
   Inception-v1, Image recognition
   Inception-v2, Image recognition
61
   ResNet-50, Image recognition
62
   Mobilenet, Image recognition
63 64
   Shufflenet, Image recognition
   SqueezeNet, Image recognition
65
   VGG-16, Image recognition
66 67 68 69 70 71
   ZFNet-512, Image recognition
   MNIST, Image recognition
   Emotion-FERPlus, Image recognition
   BVLC AlexNet, Image recognition
   BVLC GoogleNet, Image recognition
   BVLC CaffeNet, Image recognition
72 73
   BVLC R-CNN ILSVRC13, Object detection
   ArcFace, Face Detection and Recognition
74 75


76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109
.. _gpu_tensorflow:

GPU TensorFlow
==============

.. csv-table::
   :header: "TensorFlow Workload", "Genre of Deep Learning"
   :escape: ~


   Resnet50 v2, Image recognition 
   Inception V3, Image recognition
   Inception V4, Image recognition
   Inception-ResNetv2, Image recognition
   VGG-16, Image recognition 


.. _gpu_onnx:

GPU ONNX
========

.. csv-table::
   :header: "ONNX Workload", "Genre of Deep Learning"
   :escape: ~

   Inception V1, Image recognition 
   Inception V2, Image recognition 
   ResNet-50, Image recognition 
   SqueezeNet, Image recognition 
   



110 111
.. important:: Please see Intel's `Optimization Notice`_ for details on disclaimers. 

L.S. Cook's avatar
L.S. Cook committed
112
.. rubric:: Footnotes
113

Leona C's avatar
Leona C committed
114
.. [#f1] Benchmarking performance of DL systems is a young discipline; it is a
L.S. Cook's avatar
L.S. Cook committed
115 116 117 118 119
   good idea to be vigilant for results based on atypical distortions in the 
   configuration parameters. Every topology is different, and performance 
   changes can be attributed to multiple causes. Also watch out for the word 
   "theoretical" in comparisons; actual performance should not be compared to 
   theoretical performance.
120 121 122 123




L.S. Cook's avatar
L.S. Cook committed
124 125
.. _Optimization Notice: https://software.intel.com/en-us/articles/optimization-notice
.. _nGraph ONNX importer: https://github.com/NervanaSystems/ngraph-onnx/blob/master/README.md
126
.. _ONNX Model Zoo: https://github.com/onnx/models
L.S. Cook's avatar
L.S. Cook committed
127 128 129 130 131 132 133 134 135 136

.. Notice revision #20110804: Intel's compilers may or may not optimize to the same degree for 
   non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. 
   These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. 
   Intel does not guarantee the availability, functionality, or effectiveness of any optimization 
   on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this 
   product are intended for use with Intel microprocessors. Certain optimizations not specific 
   to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the 
   applicable product User and Reference Guides for more information regarding the specific 
   instruction sets covered by this notice.