Differences
This shows you the differences between two versions of the page.
Both sides previous revision Previous revision Next revision | Previous revision Next revision Both sides next revision | ||
gpu [2016/12/02 15:36] kocmanek |
gpu [2017/07/17 08:24] kocmanek [Performance tests] |
||
---|---|---|---|
Line 5: | Line 5: | ||
===== Servers with GPU units ===== | ===== Servers with GPU units ===== | ||
- | | machine | GPU; [[https:// | + | | machine |
- | | titan-gpu | + | | titan | GeForce GTX 1080 Ti; cc6.1 | 1 |
- | | twister1 | + | | titan-gpu |
- | | twister2 | + | | twister1; |
- | | kronos-dev | Tesla K40c; cc3.5 | + | | iridium |
- | | iridium | Quadro K2000; cc3.0 | 1 | 2 GB | + | | victoria; arc |
- | | victoria | + | | athena |
- | | arc | + | | dll1; dll2 | GeForce GTX 1080; cc6.1 | 8 | 8 GB each core |
- | | athena | + | | dll3; dll4; dll5 | GeForce GTX 1080 Ti; cc6.1 | 10 | 11 GB each core | | |
not used at the moment: GeForce GTX 570 (from twister2) | not used at the moment: GeForce GTX 570 (from twister2) | ||
+ | All machines have CUDA8.0 and should support both Theano and TensorFlow. | ||
Summary of future plans: | Summary of future plans: | ||
Line 24: | Line 25: | ||
* Yes, there are grant applications under review which include rack machines with GPUs, e.g. 5x2 or something like that; more will be known in 2017. | * Yes, there are grant applications under review which include rack machines with GPUs, e.g. 5x2 or something like that; more will be known in 2017. | ||
- | |||
- | | ||
- | |||
- | === Performance tests === | ||
- | |||
- | * [[http:// | ||
- | |||
- | In the following table is the experiment conducted by Tom Kocmi. You can replicate experiment: / | ||
- | |||
- | | machine | GPU; [[https:// | ||
- | | titan-gpu | ||
- | | twister1 | ||
- | | twister2 | ||
- | | kronos-dev | Tesla K40c; cc3.5 | ||
- | | iridium | Quadro K2000; cc3.0 | | | | ||
- | | victoria | ||
- | | arc | GeForce GT 630; cc3.0 | ||
- | | athena | ||
=== Individual acquisitions: | === Individual acquisitions: | ||
- | There is an easy way to get one high-end GPU: [[https:// | + | There is an easy way to get one high-end GPU: [[https:// |
Take care, however, to coordinate the grant applications a little, so that not too many arrive from UFAL within a short time: these grants are explicitly //not// intended to build GPU clusters, they are " | Take care, however, to coordinate the grant applications a little, so that not too many arrive from UFAL within a short time: these grants are explicitly //not// intended to build GPU clusters, they are " | ||
Line 53: | Line 36: | ||
* Ondřej Plátek - granted (2015) | * Ondřej Plátek - granted (2015) | ||
* Jan Hajič jr. - granted (early 2016) | * Jan Hajič jr. - granted (early 2016) | ||
- | * Jindra Helcl - planning to apply (fall 2016) | ||
- | ===== Installed toolkits ===== | ||
- | //This should mention where each interesting toolkit lives (on a particular machine).// | + | |
- | ==== TensorFlow | + | ===== How to use cluster ===== |
- | [[https:// | + | In this section will be explained how to use cluster properly. |
+ | ==== TensorFlow | ||
- | OP: I created [[https:// | + | Majority people at UFAL use TensorFlow. To start using it you need to create python virtual environment (virtualenv or use Anaconda for it). Into the environment you must place TensorFlow. The TF is either |
- | === Select GPU device === | + | pip install tensorflow |
+ | pip install tensorflow-gpu | ||
+ | |||
+ | You can use prepared environment by adding into your ~/.bashrc | ||
- | Use variable CUDA_VISIBLE_DEVICES to constrain tensorflow to compute only on the selected one. For the use of first GPU use: | + | |
- | < | + | |
- | To list available devices, use: | + | And then you can activate your environment: |
- | < | + | |
- | ===== Basic commands | + | source activate tf1 |
+ | source activate tf1cpu | ||
+ | |||
+ | This environment have TensorFlow 1.0 and all necessary requirements for NeuralMonkey. | ||
+ | |||
+ | ==== Using cluster ==== | ||
+ | |||
+ | Rule number one, always use the GPU queue (never log in machine by ssh). Always use qsub or qsubmit with proper arguments. | ||
+ | |||
+ | For testing and using the cluster interactively you can use qrsh (this should not be used for long running experiments since the console is not closed on the end of the experiment). Following command will assign you a GPU and creates interactive console. | ||
+ | |||
+ | qrsh -q gpu.q -l gpu=1, | ||
+ | |||
+ | For running experiments you must use qsub command: | ||
+ | |||
+ | qsub -q gpu.q -l gpu=1, | ||
+ | |||
+ | Cleaner way to use cluster is with / | ||
+ | |||
+ | qsubmit --gpumem=2G --queue=" | ||
+ | |||
+ | It is recommended to use priority -100 if you are not rushing for the results and don't need to leap over your colleagues jobs. | ||
+ | ==== Basic commands ==== | ||
lspci | lspci | ||
Line 90: | Line 95: | ||
/ | / | ||
# shows CUDA capability etc. | # shows CUDA capability etc. | ||
+ | | ||
+ | === Select GPU device === | ||
+ | |||
+ | Use variable CUDA_VISIBLE_DEVICES to constrain tensorflow to compute only on the selected one. For the use of first GPU use (GPU queue do this for you): | ||
+ | export CUDA_VISIBLE_DEVICES=0 | ||
+ | |||
+ | To list available devices, use: | ||
+ | / | ||
+ | |||
+ | ===== Performance tests ===== | ||
+ | |||
+ | * [[http:// | ||
+ | |||
+ | In the following table is the experiment conducted by Tom Kocmi. You can replicate experiment: / | ||
+ | |||
+ | | machine | Setup; CPU/GPU; [[https:// | ||
+ | | athena | ||
+ | | dll2 | (2 GPU) GeForce GTX 1080; cc6.1 | | ||
+ | | titan | GeForce GTX 1080 Ti | | ||
+ | | dll1 | (2 GPU) GeForce GTX 1080; cc6.1 | | ||
+ | | dll2 | (2 GPU) GeForce GTX 1080; cc6.1 | | ||
+ | | titan-gpu | ||
+ | | kronos-dev | Tesla K40c; cc3.5 | | ||
+ | | twister2 | ||
+ | | twister1 | ||
+ | | helena1 | ||
+ | | belzebub | ||
+ | | iridium | ||
+ | | helena7 | ||
+ | | arc | GeForce GT 630; cc3.0 | 103:42:30 | (approximated after 66 hours) | | ||
+ | | lucifer4 | ||
+ | |||
+ | |||
+ | === Second Benchmark === | ||
+ | |||
+ | The previous benchmark only compares the speed of processing units within the GPUs and do not take into account the size of memory. Therefore I have conducted another benchmark, this time for each graphic card I have increased the batch size as much as possible so the model still could fit into the GPU (the previous benchmark had batch size 20). This way the results should be more representative of the power for each GPU. | ||
+ | |||
+ | | GPU; Cuda capability | ||
+ | | Tesla K40c; cc3.5 | 12 GB | | ||
+ | | GeForce GTX 1080 Ti; cc6.1 | 11 GB | 00:55:56 | 2300 | dll5 | | ||
+ | | GeForce GTX 1080; cc6.1 | 8 GB | 01:10:57 | 1700 | dll1 | | ||
+ | | GeForce GTX 1080; cc6.1 | 8 GB | | ||
+ | | GeForce GTX Titan Z; cc3.5 | 6 GB | 02:20:47 | 1100 | titan-gpu | | ||
+ | | Quadro K2000; cc3.0 | 2 GB | 28:15:26 | 50 | iridium | | ||
===== Links ===== | ===== Links ===== |