mirror of
https://github.com/NVIDIA/cuda-samples.git
synced 2024-10-06 22:19:17 +08:00
0bce11f7a5
-- remove MDCG support on windows -- integrate fixes to nvrtc samples makefile and driver api samples makefiles
65 lines
2.5 KiB
Markdown
65 lines
2.5 KiB
Markdown
# conjugateGradientMultiDeviceCG - conjugateGradient using MultiDevice Cooperative Groups
|
|
|
|
## Description
|
|
|
|
This sample implements a conjugate gradient solver on multiple GPUs using Multi Device Cooperative Groups, also uses Unified Memory optimized using prefetching and usage hints.
|
|
|
|
## Key Concepts
|
|
|
|
Unified Memory, Linear Algebra, Cooperative Groups, MultiDevice Cooperative Groups
|
|
|
|
## Supported SM Architectures
|
|
|
|
[SM 6.0 ](https://developer.nvidia.com/cuda-gpus) [SM 6.1 ](https://developer.nvidia.com/cuda-gpus) [SM 7.0 ](https://developer.nvidia.com/cuda-gpus)
|
|
|
|
## Supported OSes
|
|
|
|
Linux
|
|
|
|
## Supported CPU Architecture
|
|
|
|
x86_64, ppc64le
|
|
|
|
## CUDA APIs involved
|
|
|
|
### [CUDA Runtime API](http://docs.nvidia.com/cuda/cuda-runtime-api/index.html)
|
|
cudaMemAdvise, cudaMemPrefetchAsync, cudaLaunchCooperativeKernelMultiDevice, cudaStreamSynchronize, cudaOccupancyMaxActiveBlocksPerMultiprocessor
|
|
|
|
## Dependencies needed to build/run
|
|
[UVM](../../README.md#uvm), [MDCG](../../README.md#mdcg)
|
|
|
|
## Prerequisites
|
|
|
|
Download and install the [CUDA Toolkit 9.2](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
|
|
Make sure the dependencies mentioned in [Dependencies]() section above are installed.
|
|
|
|
## Build and Run
|
|
|
|
### Linux
|
|
The Linux samples are built using makefiles. To use the makefiles, change the current directory to the sample directory you wish to build, and run make:
|
|
```
|
|
$ cd <sample_dir>
|
|
$ make
|
|
```
|
|
The samples makefiles can take advantage of certain options:
|
|
* **TARGET_ARCH=<arch>** - cross-compile targeting a specific architecture. Allowed architectures are x86_64, ppc64le.
|
|
By default, TARGET_ARCH is set to HOST_ARCH. On a x86_64 machine, not setting TARGET_ARCH is the equivalent of setting TARGET_ARCH=x86_64.<br/>
|
|
`$ make TARGET_ARCH=x86_64` <br/> `$ make TARGET_ARCH=ppc64le` <br/>
|
|
See [here](http://docs.nvidia.com/cuda/cuda-samples/index.html#cross-samples) for more details.
|
|
* **dbg=1** - build with debug symbols
|
|
```
|
|
$ make dbg=1
|
|
```
|
|
* **SMS="A B ..."** - override the SM architectures for which the sample will be built, where `"A B ..."` is a space-delimited list of SM architectures. For example, to generate SASS for SM 50 and SM 60, use `SMS="50 60"`.
|
|
```
|
|
$ make SMS="50 60"
|
|
```
|
|
|
|
* **HOST_COMPILER=<host_compiler>** - override the default g++ host compiler. See the [Linux Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html#system-requirements) for a list of supported host compilers.
|
|
```
|
|
$ make HOST_COMPILER=g++
|
|
```
|
|
|
|
## References (for more details)
|
|
|