While getting started with CUDA on Windows or on WSL (same on Linux) requires to install some stuff, it is not the case when using Google Colab. For those who don't know, Google Colab is a hosted jupyter notebook service which requires no setup and where you can access resources like CPU, GPU and TPU for free (you can also get some subscriptions for more GPU power, etc... but free tier is good enough for training and learning).
Let's also check if the CUDA Compiler is ready:
Now it would be nice to run the same deviceQuery sample script as the one we ran on Windows and on WSL. As a reminder, deviceQuery is a sample script provided by NVidia in this git repo.
To achieve that, we need to:
Yeah OK that's boring and it looks complicated so I got you covered; just use that notebook I made for you and click on play (or CTRL + F9 will run all the cells at once !).
In Colab:
And here is the deviceQuery CUDA sample script output on Google Colab using that Tesla T4 GPU:
- First thing is to head over to Google Colab and create a new notebook
- Go to Runtime => Change runtype type and choose GPU
- Click Connect on the top right and you'll get your notebook connected some RAM, some disk and a GPU
!nvidia-smi
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.104.05 Driver Version: 535.104.05 CUDA Version: 12.2 |
|-----------------------------------------+----------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+======================+======================|
| 0 Tesla T4 Off | 00000000:00:04.0 Off | 0 |
| N/A 39C P8 9W / 70W | 0MiB / 15360MiB | 0% Default |
| | | N/A |
+-----------------------------------------+----------------------+----------------------+
+---------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=======================================================================================|
| No running processes found |
+---------------------------------------------------------------------------------------+
We got assigned a Tesla T4 GPU which is not a top notch GPU but it is faaaaaaaaar more than what we would ever need for training/learning (and remember, all of this is free; Google's treat).Let's also check if the CUDA Compiler is ready:
!nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2023 NVIDIA Corporation
Built on Tue_Aug_15_22:02:13_PDT_2023
Cuda compilation tools, release 12.2, V12.2.140
Build cuda_12.2.r12.2/compiler.33191640_0
Great, all good in just a few clicks!Now it would be nice to run the same deviceQuery sample script as the one we ran on Windows and on WSL. As a reminder, deviceQuery is a sample script provided by NVidia in this git repo.
To achieve that, we need to:
- Get the C++ source of deviceQuery
- Get the source of helper_cuda.h
- Get the source of helper_string.h
- Compile deviceQuery
- Run it
Yeah OK that's boring and it looks complicated so I got you covered; just use that notebook I made for you and click on play (or CTRL + F9 will run all the cells at once !).
In Colab:
- File => Open Notebook(CTRL + O)
- Choose Github
- Paste the link of the notebook: https://github.com/freddenis/cuda-training/blob/main/00_cuda_deviceQuery.ipynb
- Run cell by cell to see what happens or CTRL + F9 and check the outoput the bottom of the page
And here is the deviceQuery CUDA sample script output on Google Colab using that Tesla T4 GPU:
!time ./deviceQuery ./deviceQuery Starting... CUDA Device Query (Runtime API) version (CUDART static linking) Detected 1 CUDA Capable device(s) Device 0: "Tesla T4" CUDA Driver Version / Runtime Version 12.2 / 12.2 CUDA Capability Major/Minor version number: 7.5 Total amount of global memory: 15102 MBytes (15835660288 bytes) (040) Multiprocessors, (064) CUDA Cores/MP: 2560 CUDA Cores <== This is still a beast of a GPU! GPU Max Clock rate: 1590 MHz (1.59 GHz) Memory Clock rate: 5001 Mhz Memory Bus Width: 256-bit L2 Cache Size: 4194304 bytes Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384) Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers Total amount of constant memory: 65536 bytes Total amount of shared memory per block: 49152 bytes Total shared memory per multiprocessor: 65536 bytes Total number of registers available per block: 65536 Warp size: 32 Maximum number of threads per multiprocessor: 1024 Maximum number of threads per block: 1024 Max dimension size of a thread block (x,y,z): (1024, 1024, 64) Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535) Maximum memory pitch: 2147483647 bytes Texture alignment: 512 bytes Concurrent copy and kernel execution: Yes with 3 copy engine(s) Run time limit on kernels: No Integrated GPU sharing Host Memory: No Support host page-locked memory mapping: Yes Alignment requirement for Surfaces: Yes Device has ECC support: Enabled Device supports Unified Addressing (UVA): Yes Device supports Managed Memory: Yes Device supports Compute Preemption: Yes Supports Cooperative Kernel Launch: Yes Supports MultiDevice Co-op Kernel Launch: Yes Device PCI Domain ID / Bus ID / location ID: 0 / 0 / 4 Compute Mode: < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) > deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.2, CUDA Runtime Version = 12.2, NumDevs = 1 Result = PASS real 0m0.133s user 0m0.014s sys 0m0.114sWhat an easy and very powerful tool! (ah yes, it is also free)
