AI Benchmark Results

Here we show the data and performance of GPUs by AI-Benchmark on ROScube series.

GPU Burn in / Test by AI-Benchmark.

AI Benchmark Alpha is an open source python library for evaluating AI performance of various hardware platforms, including CPUs, GPUs and TPUs.

It can be downloaded to any system running Windows, Linux or macOS.

In total, AI Benchmark consists of 42 tests and 19 sections provided below:

• MobileNet-V2 [classification]

• Inception-V3 [classification]

• Inception-V4 [classification]

• Inception-ResNet-V2 [classification]

• ResNet-V2-50 [classification]

• ResNet-V2-152 [classification]

• VGG-16 [classification]

• SRCNN 9-5-5 [image-to-image mapping]

• VGG-19 [image-to-image mapping]

• ResNet-SRGAN [image-to-image mapping]

• ResNet-DPED [image-to-image mapping]

• U-Net [image-to-image mapping]

• Nvidia-SPADE [image-to-image mapping]

• ICNet [image segmentation]

• PSPNet [image segmentation]

• DeepLab [image segmentation]

• Pixel-RNN [inpainting]

• LSTM [sentence sentiment analysis]

• GNMT [text translation]

In the testing, we can get the training time of different models.

Then AI-Benchmark will provide the score which the performance of GPUs.

GPU Score

After testing, we can go the RANK to compare the performance of GPUs.

In the rank, you can see the open testing data of different GPUs.

It is a fairer way to evaluate the performance of GPUs.

Here, we provided some test data of GPUs which we actual tested on ROScube series.

Device	GPU	Inference Score	Training Score	AI Score
ROScube-I-E	NVIDIA RTX A2000 12G	8530	8915	17445
ROScube-I-E	NVIDIA GTX 1050 2G	3200	X	X
ROScube-X-580	Jetson AGX Xavier	2142	2192	4334
ROScube-Pico-NX	Jetson Xavier NX	709	787	1496

Note

GPU with at least 2GB of RAM is required for running inference tests / 4GB of RAM for training tests.

The benchmark is compatible with both TensorFlow 1.x and 2.x versions.