phaniarnab commented on a change in pull request #1269:
URL: https://github.com/apache/systemds/pull/1269#discussion_r628793336
##########
File path: docs/site/gpu.md
##########
@@ -0,0 +1,108 @@
+---
+layout: site
+title: Run SystemDS with GPU
+---
+<!--
+{% comment %}
+Licensed to the Apache Software Foundation (ASF) under one or more
+contributor license agreements. See the NOTICE file distributed with
+this work for additional information regarding copyright ownership.
+The ASF licenses this file to you under the Apache License, Version 2.0
+(the "License"); you may not use this file except in compliance with
+the License. You may obtain a copy of the License at
+
+http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+{% endcomment %}
+-->
+
+This guide helps in installing pre-requisites for running with GPU and run
instructions.
+
+- [Requirements](#requirements)
+- [Command-line users](#command-line-users)
+- [Scala Users](#scala-users)
+- [Advanced Configuration](#advanced-configuration)
+ - [Using single precision](#using-single-precision)
+- [Training very deep network](#training-very-deep-network)
+ - [Shadow buffer](#shadow-buffer)
+ - [Unified memory allocator](#unified-memory-allocator)
+
+## Requirements
+
+### Hardware
+
+The following GPUs are supported:
+
+* NVIDIA GPU cards with CUDA architectures 3.5, 5.0, 6.0, 7.0, 7.5, 8.0 and
higher than 8.0.
+For CUDA enabled gpu cards at [CUDA
GPUS](https://developer.nvidia.com/cuda-gpus)
+
+* For GPUs with unsupported CUDA architectures, or to avoid JIT compilation
from PTX, or to
+use difference versions of the NVIDIA libraries, see the Linux build from
source guide.
+
+* Release artifacts contain PTX code for the latest supported CUDA
architecture.
+
+### Software
+
+The following NVIDIA software is required to be installed in your system:
+
+CUDA toolkit
+
+ 1. [NVIDIA GPU drivers](https://www.nvidia.com/drivers) - CUDA 10.2 requires
>= 440.33 driver. see
+ [CUDA
compatibility](https://docs.nvidia.com/deploy/cuda-compatibility/index.html).
+ 3. [CUDA 10.2](https://developer.nvidia.com/cuda-10.2-download-archive)
+ 4. [CUDNN 7.x](https://developer.nvidia.com/cudnn)
+
+
+## Command-line users
+
+To enable the GPU backend via command-line, please provide
`systemds-*-extra.jar` in the classpath and `-gpu` flag.
+
+```
+spark-submit --jars systemds-*-extra.jar SystemDS.jar -f myDML.dml -gpu
+```
+
+To skip memory-checking and force all GPU-enabled operations on the GPU,
please provide `force` option to the `-gpu` flag.
+
+```
+spark-submit --jars systemds-*-extra.jar SystemDS.jar -f myDML.dml -gpu force
+```
+
+## Scala users
+
+To enable the GPU backend via command-line, please provide
`systemds-*-extra.jar` in the classpath and use
+the `setGPU(True)` method of MLContext API to enable the GPU usage.
+
+```
+spark-shell --jars systemds-*-extra.jar,SystemDS.jar
+```
+
+## Advanced Configuration
+
+### Using single precision
+
+By default, SystemDS uses double precision to store its matrices in the GPU
memory.
+To use single precision, the user needs to set the configuration property
`sysds.floating.point.precision`
+to `single`. However, with exception of BLAS operations, SystemDS always
performs all CPU operations
+in double precision.
+
+### Training very deep network
+
+#### Shadow buffer
+To train very deep network with double precision, no additional configurations
are necessary.
+But to train very deep network with single precision, the user can speed up
the eviction by
+using shadow buffer. The fraction of the driver memory to be allocated to the
shadow buffer can
+be set by using the configuration property
`sysds.gpu.eviction.shadow.bufferSize`.
+In the current version, the shadow buffer is currently not guarded by SystemDS
+and can potentially lead to OOM if the network is deep as well as wide.
+
+#### Unified memory allocator
+
+By default, SystemDS uses CUDA's memory allocator and performs on-demand
eviction
+using the eviction policy set by the configuration property
`sysds.gpu.eviction.policy`.
+To use CUDA's unified memory allocator that performs page-level eviction
instead,
+please set the configuration property `sysml.gpu.memory.allocator` to
`unified_memory`.
Review comment:
Currently, the eviction policy is always LRU regardless of what is set
in sysds.gpu.eviction.policy. Please reword accordingly.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
