SABERTOOTH cuDNN Tutorial
| Complexity: | Low |
|---|---|
| Payloads: | SABERTOOTH |
| Windows: | PAYLOAD_SABERTOOTH |
This tutorial demonstrates CuDNN v7.5.0 by running the samples that come shipped with it.
Prerequisites
All tutorials require the steps outlined in the Getting Started Guide.
Overview
The tutorial comes with 2 scripts:
cudnn_demo- runs in-orbit on the SABERTOOTH to demonstrate cuDNN runningdeploy- run by the user on the ground to uploadcudnn_demoand schedule it to execute in aPAYLOAD_SABERTOOTHwindow
cudnn_demo Script
cudnn_demo builds & runs the following cuDNN samples:
mnistCUDNNRNN
First, the CUDA compiler is checked for availability:
$ /usr/local/cuda/bin/nvcc -V
Next, the cuDNN samples source is copied to ~/ as the system location is not writable:
$ cp -a /usr/src/cudnn_samples_v7 ~/
Finally some of the CUDA samples are built & run. Here are the commands for building & running mnistCUDNN:
$ cd ~/cudnn_samples_v7/mnistCUDNN
$ make
$ ./mnistCUDNN
Example Output
Linking agains cublasLt = false CUDA VERSION: 10000 TARGET ARCH: aarch64 HOST_ARCH: aarch64 TARGET OS: linux SMS: 30 35 50 53 60 61 62 70 72 75 /usr/local/cuda/bin/nvcc -ccbin g++ -I/usr/local/cuda/include -IFreeImage/include -m64 -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=sm_50 -gencode arch=compute_53,code=sm_53 -gencode arch=compute_60,code=sm_60 -gencode arch=compute_61,code=sm_61 -gencode arch=compute_62,code=sm_62 -gencode arch=compute_70,code=sm_70 -gencode arch=compute_72,code=sm_72 -gencode arch=compute_75,code=sm_75 -gencode arch=compute_75,code=compute_75 -o fp16_dev.o -c fp16_dev.cu g++ -I/usr/local/cuda/include -IFreeImage/include -o fp16_emu.o -c fp16_emu.cpp g++ -I/usr/local/cuda/include -IFreeImage/include -o mnistCUDNN.o -c mnistCUDNN.cpp /usr/local/cuda/bin/nvcc -ccbin g++ -m64 -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=sm_50 -gencode arch=compute_53,code=sm_53 -gencode arch=compute_60,code=sm_60 -gencode arch=compute_61,code=sm_61 -gencode arch=compute_62,code=sm_62 -gencode arch=compute_70,code=sm_70 -gencode arch=compute_72,code=sm_72 -gencode arch=compute_75,code=sm_75 -gencode arch=compute_75,code=compute_75 -o mnistCUDNN fp16_dev.o fp16_emu.o mnistCUDNN.o -I/usr/local/cuda/include -IFreeImage/include -LFreeImage/lib/linux/aarch64 -LFreeImage/lib/linux -lcudart -lcublas -lcudnn -lfreeimage -lstdc++ -lm FreeImage/lib/linux/aarch64/libfreeimage.a(strenc.o): In function `StrIOEncInit': strenc.c:(.text+0x1294): warning: the use of `tmpnam' is dangerous, better use `mkstemp' cudnnGetVersion() : 7500 , CUDNN_VERSION from cudnn.h : 7500 (7.5.0) Host compiler version : GCC 7.4.0 There are 1 CUDA capable devices on your machine : device 0 : sms 1 Capabilities 5.3, SmClock 921.6 Mhz, MemSize (Mb) 1980, MemClock 12.8 Mhz, Ecc=0, boardGroupID=0 Using device 0 Testing single precision Loading image data/one_28x28.pgm Performing forward propagation ... Testing cudnnGetConvolutionForwardAlgorithm ... Fastest algorithm is Algo 1 Testing cudnnFindConvolutionForwardAlgorithm ... ^^^^ CUDNN_STATUS_SUCCESS for Algo 0: 0.251667 time requiring 0 memory ^^^^ CUDNN_STATUS_SUCCESS for Algo 2: 0.382813 time requiring 57600 memory ^^^^ CUDNN_STATUS_SUCCESS for Algo 4: 2.635729 time requiring 207360 memory ^^^^ CUDNN_STATUS_SUCCESS for Algo 5: 12.472500 time requiring 203008 memory ^^^^ CUDNN_STATUS_SUCCESS for Algo 7: 12.811354 time requiring 2057744 memory Resulting weights from Softmax: 0.0000000 0.9999399 0.0000000 0.0000000 0.0000561 0.0000000 0.0000012 0.0000017 0.0000010 0.0000000 Loading image data/three_28x28.pgm Performing forward propagation ... Resulting weights from Softmax: 0.0000000 0.0000000 0.0000000 0.9999288 0.0000000 0.0000711 0.0000000 0.0000000 0.0000000 0.0000000 Loading image data/five_28x28.pgm Performing forward propagation ... Resulting weights from Softmax: 0.0000000 0.0000008 0.0000000 0.0000002 0.0000000 0.9999820 0.0000154 0.0000000 0.0000012 0.0000006 Result of classification: 1 3 5 Test passed! Testing half precision (math in single precision) Loading image data/one_28x28.pgm Performing forward propagation ... Testing cudnnGetConvolutionForwardAlgorithm ... Fastest algorithm is Algo 1 Testing cudnnFindConvolutionForwardAlgorithm ... ^^^^ CUDNN_STATUS_SUCCESS for Algo 1: 0.171354 time requiring 3464 memory ^^^^ CUDNN_STATUS_SUCCESS for Algo 0: 0.206354 time requiring 0 memory ^^^^ CUDNN_STATUS_SUCCESS for Algo 2: 0.319636 time requiring 28800 memory ^^^^ CUDNN_STATUS_SUCCESS for Algo 4: 2.538021 time requiring 207360 memory ^^^^ CUDNN_STATUS_SUCCESS for Algo 5: 12.519062 time requiring 203008 memory Resulting weights from Softmax: 0.0000001 1.0000000 0.0000001 0.0000000 0.0000563 0.0000001 0.0000012 0.0000017 0.0000010 0.0000001 Loading image data/three_28x28.pgm Performing forward propagation ... Resulting weights from Softmax: 0.0000000 0.0000000 0.0000000 1.0000000 0.0000000 0.0000714 0.0000000 0.0000000 0.0000000 0.0000000 Loading image data/five_28x28.pgm Performing forward propagation ... Resulting weights from Softmax: 0.0000000 0.0000008 0.0000000 0.0000002 0.0000000 1.0000000 0.0000154 0.0000000 0.0000012 0.0000006 Result of classification: 1 3 5 Test passed!
Upload & Deploy
Run deploy. The script uploads cudnn_demo and schedules it's execution in 24 hours.
$ ./deploy "[YOUR_AUTH_TOKEN]" [YOUR_SAT_ID]
Review
Once the window has completed and enough time has passed for the log to download, it can be reviewed in AWS S3 (see Hello World tutorial).
Next Steps
This is the last tutorial. Next step is to better understand the development environments available.