[{"child":[],"type":"tab","idName":"","className":"col-4","data":[{"link":"","text":"Benefits","blank":false},{"link":"function","title":"","blank":false,"text":"Features"},{"link":"price","title":"","blank":false,"text":"Pricing"},{"link":"${guideUrl}/en/tensorflow-tensorflow-2-1","title":"","blank":true,"text":"Guide"}],"uId":"15699075245756","active":false},{"child":[{"child":[],"type":"text","text":"

Available in Classic

\n","useYn":"Y","active":false,"uId":"160026562827330"},{"child":[],"type":"title","idName":"","className":"","text":"

Easily create a TensorFlow cluster

\nYou can easily set up a TensorFlow distributed parallel processing environment with the CLI to increase the efficiency of running TensorFlow with a large batch size.

","uId":"15687730540678","active":false,"useYn":"Y"}],"type":"section","idName":"","className":"","uId":"15687730456665","active":false},{"child":[{"child":[],"type":"text","idName":"","className":"","text":"CLI-based configuration\n: Once you create a master server, a VM server for managing clusters, run CLI commands to get a pre-configured server node with libraries installed on it. You can freely add or delete worker and parameter server nodes to manage cluster configurations.","uId":"15687731992269","active":false,"useYn":"Y"},{"child":[],"type":"text","idName":"","className":"","text":"Efficient data deployment and execution\n: You can create, delete, and manage an NAS volume with the CLI commands provided by default. You can also mount and share both training code and data at the same time between server nodes for easy deployment and execution.","uId":"156897102242030","active":false,"useYn":"Y"},{"child":[],"type":"text","idName":"","className":"","text":"Minimal code changes\n: You only need to make minimal changes to your TensorFlow code as you can get the data of the pre-configured cluster from your code.","uId":"15699967122995","active":false,"useYn":"Y"},{"child":[],"type":"text","idName":"","className":"","text":"Support for Intel Math Kernel Library (MKL)\n: The cluster comes with a version of TensorFlow that seeks to enhance performance for deep learning and machine learning on Intel Xeon processors using the Intel Math Kernel Library (MKL). The result is more than 200 percent improvement in performance in the workloads.","uId":"156897102226129","active":false,"useYn":"Y"}],"type":"section","idName":"","className":"def","uId":"15687730479636","active":false},{"child":[{"child":[],"type":"text","text":"### Features\n","uId":"156878414867311","active":false,"useYn":"Y"},{"child":[],"type":"text","text":"#### Operate server nodes in numerous ways\n\nYou can create server nodes in a cluster simply by running CLI commands on a master server, and stop or restart the nodes all at once or individually without having to access NAVER Cloud Platform's web console. You can add, delete, or return some of the server nodes.","uId":"156897110902332","active":false,"useYn":"Y"},{"child":[],"type":"text","text":"#### Easy deployment using shared storage\n\nYou can create and expand NAS storage simply by running CLI commands on a master server without having to access NAVER Cloud Platform's web console. You can mount or unmount NAS storage to all of the nodes simultaneously with a one-time command, enabling you to protect your codes and training data. At the same time, you can reduce dependency on clusters while solving problems with deploying and sharing codes and data. By default, you can create 500 GB of NAS storage and expand as necessary to accommodate terabytes of training data.","uId":"156897110854331","active":false,"useYn":"Y"},{"child":[],"type":"text","text":"#### Submit jobs to a cluster\n\nOnce a change has been committed to a master server, it is applied automatically to server nodes. This allows you to run TensorFlow Cluster with minimal changes to the codes you were previously running on a single server (VM). After all jobs on the job server are complete, the parameter server stops responding automatically, so that no separate cleanup process is required when restarting the process.","uId":"156897115198033","active":false,"useYn":"Y"},{"child":[],"type":"text","text":"#### View job logs for a cluster\n\nThe jobs in a cluster run in the background and logs are redirected to the master server for integration. You can view logs of the server nodes in real time using the CLI monitor command and a list of previously executed tasks using the CLI history command.","uId":"156897117583735","active":false,"useYn":"Y"},{"child":[],"type":"text","text":"#### A broad spectrum of server node specifications\n\nChoose from the below five server specifications provided by default (the GPU server will be available soon):","uId":"156897117538234","active":false,"useYn":"Y"},{"child":[],"type":"custom","idName":"","className":"","text":"

\n MINI (4 vCPUs, 16 GB memory, 50 GB HDD): suited for testing clusters or handling small-sized workloads\n

BASIC (8 vCPUs, 32 GB memory, 50 GB HDD): suited for handling medium-sized workloads

HIGH (16 vCPUs, 32 GB memory, 50 GB HDD): suited for handling large-sized workloads

\n GPU1 (1 GPU, 24 GB GPU memory, 4 vCPUs, 30 GB memory, 50 GB SSD) – Scale a single GPU proportional to the number of\n cluster nodes\n

\n GPU2 (2 GPUs, 48 GB GPU memory, 8 vCPUs, 60 GB memory, 50 GB SSD) – Scale a dual GPU proportional to the number of cluster nodes (suited for handling very large workloads)
(Note that node specifications are configured identically and all of the nodes are recognized as worker server nodes. You can specify the number of nodes for parameter servers.)\n

\n","uId":"156897120316036","active":false,"useYn":"Y"},{"child":[],"type":"text","text":"#### Note\n\nNAVER Cloud’s TensorFlow Cluster uses TensorFlow, an open-source machine learning software library developed by the Google Brain team.","uId":"156878428729315","active":false,"useYn":"Y"}],"type":"section","idName":"","className":"function","uId":"15687730494187","active":false},{"child":[{"child":[],"type":"custom","idName":"","className":"","text":"

","uId":"156897131994337","active":false},{"child":[],"type":"text","text":"### Pricing Information\n\nNo additional fee is charged on TensorFlow Cluster except for the server fee to configure the cluster. \n(However, stop the server nodes to avoid excessive charges in case you are not submitting the cluster job.)","uId":"156878443416316","active":false},{"child":[],"type":"table","idName":"","className":"","legend":"","text":"OSVersionUsage Fee (Month)\n Ubuntu 16.04
(GPU server nodes will be supported soon)
❖ TensorFlow is not installed in the master\n node for quick server creation. Users must install it if necessary – See the manual.\n \n Tensorflow 1.3 (Stable Latest)
(The package version distributed may change depending on the version-up\n speed of TensorFlow.)\n \n ❖ Recommended specification for cluster testing purposes or handling little workload: KRW 205,000 a month per\n server node for 4 vCPUs and Mem 16G.\n \n","uId":"156878445919017","active":false,"vat":"Y","useYn":"Y"}],"type":"section","idName":"","className":"price","uId":"156878413787310","active":false}]

AI

Hybrid

Finance

Game

Global Service

Live commerce

SMARTWORK

TensorFlow Cluster