Updated naming on code files
This commit is contained in:
parent
0788ec875f
commit
53039040ba
13 changed files with 29 additions and 29 deletions
|
|
@ -21,9 +21,9 @@ import os
|
|||
import kp
|
||||
import sphinx_material
|
||||
|
||||
project = 'Vulkan Kompute'
|
||||
project = 'Kompute'
|
||||
copyright = '2020, The Institute for Ethical AI & Machine Learning'
|
||||
html_title = "Vulkan Kompute Documentation (Python & C++)"
|
||||
html_title = "Kompute Documentation (Python & C++)"
|
||||
author = 'Alejandro Saucedo'
|
||||
|
||||
# The full version, including alpha/beta/rc tags
|
||||
|
|
@ -74,7 +74,7 @@ if html_theme == 'sphinx_material':
|
|||
'color_primary': 'red',
|
||||
'color_accent': 'light-blue',
|
||||
'repo_url': 'https://github.com/EthicalML/vulkan-kompute/',
|
||||
'repo_name': 'Vulkan Kompute',
|
||||
'repo_name': 'Kompute',
|
||||
'globaltoc_depth': 2,
|
||||
'globaltoc_collapse': False,
|
||||
'globaltoc_includehidden': False,
|
||||
|
|
|
|||
|
|
@ -1,13 +1,13 @@
|
|||
|
||||
# Kompute Mobile App Integration (Android)
|
||||
|
||||
This is the accompanying code for the Blog post ["Supercharging your Mobile Apps with On-Device GPU Accelerated Machine Learning using the Android NDK & Vulkan Kompute"](https://towardsdatascience.com/gpu-accelerated-machine-learning-in-your-mobile-applications-using-the-android-ndk-vulkan-kompute-1e9da37b7617).
|
||||
This is the accompanying code for the Blog post ["Supercharging your Mobile Apps with On-Device GPU Accelerated Machine Learning using the Android NDK & Kompute"](https://towardsdatascience.com/gpu-accelerated-machine-learning-in-your-mobile-applications-using-the-android-ndk-vulkan-kompute-1e9da37b7617).
|
||||
|
||||
<table>
|
||||
<tr>
|
||||
|
||||
<td width="70%">
|
||||
<h3>Example Running Logistic Regression with Vulkan Kompute in Android Phones</h3>
|
||||
<h3>Example Running Logistic Regression with Kompute in Android Phones</h3>
|
||||
|
||||
<p>
|
||||
This example provides an end to end example that can be run using android studio.
|
||||
|
|
|
|||
|
|
@ -34,11 +34,11 @@ In order to run this example, you will need the following dependencies:
|
|||
* REQUIRED
|
||||
+ The Vulkan SDK must be installed
|
||||
* OPTIONAL
|
||||
+ Vulkan Kompute library must be accessible (by default it uses the source directory)
|
||||
+ Kompute library must be accessible (by default it uses the source directory)
|
||||
+ SPDLOG - for logging
|
||||
+ FMT - for text formatting
|
||||
|
||||
We will cover how you can install Vulkan Kompute in the next section.
|
||||
We will cover how you can install Kompute in the next section.
|
||||
|
||||
For the Vulkan SDK, the simplest way to install it is through [their website](https://vulkan.lunarg.com/sdk/home). You just have to follow the instructions for the relevant platform.
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
# Vulkan Kompute Godot Example
|
||||
# Kompute Godot Example
|
||||
|
||||

|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
# Vulkan Kompute Godot Example
|
||||
# Kompute Godot Example
|
||||
|
||||

|
||||
|
||||
|
|
|
|||
|
|
@ -34,7 +34,7 @@ In order to run this example, you will need the following dependencies:
|
|||
* REQUIRED
|
||||
+ The Vulkan SDK must be installed
|
||||
* OPTIONAL
|
||||
+ Vulkan Kompute library must be accessible (by default it uses the source directory)
|
||||
+ Kompute library must be accessible (by default it uses the source directory)
|
||||
+ SPDLOG - for logging
|
||||
+ FMT - for text formatting
|
||||
|
||||
|
|
|
|||
|
|
@ -1,12 +1,12 @@
|
|||
# Kompute Pi4 Mesa Build Example
|
||||
|
||||
The Raspberry Pi 4 is an impressive little computer when you consider that the Broadcom GPU is able to run 2x 4K displays. This example intends to show how to get Vulkan Kompute up and running on a Raspberry Pi 4. This has huge potential for edge processing using the power of the Pi 4 GPU.
|
||||
The Raspberry Pi 4 is an impressive little computer when you consider that the Broadcom GPU is able to run 2x 4K displays. This example intends to show how to get Kompute up and running on a Raspberry Pi 4. This has huge potential for edge processing using the power of the Pi 4 GPU.
|
||||
|
||||
Special thanks to [Alejandro Piñeiro](https://blogs.igalia.com/apinheiro/) and others for their work on Broadcom drivers for [Mesa](https://gitlab.freedesktop.org/mesa/mesa) which make this example possible.
|
||||
|
||||
## Raspberry Pi Operating System
|
||||
|
||||
For this experiment we used [RaspiOS Lite 2021-01-12](https://downloads.raspberrypi.org/raspios_lite_armhf/images/raspios_lite_armhf-2021-01-12/2021-01-11-raspios-buster-armhf-lite.zip), though it is likely best to start with the latest available operation system from [Raspberry Pi operating system images](https://www.raspberrypi.org/software/operating-systems/). In other experiments the full Raspberry Pi operating system (with desktop environment) was found to work. However, when attempting to use Ubuntu on the Raspberry Pi we were not able to run the Python Vulkan Kompute examples.
|
||||
For this experiment we used [RaspiOS Lite 2021-01-12](https://downloads.raspberrypi.org/raspios_lite_armhf/images/raspios_lite_armhf-2021-01-12/2021-01-11-raspios-buster-armhf-lite.zip), though it is likely best to start with the latest available operation system from [Raspberry Pi operating system images](https://www.raspberrypi.org/software/operating-systems/). In other experiments the full Raspberry Pi operating system (with desktop environment) was found to work. However, when attempting to use Ubuntu on the Raspberry Pi we were not able to run the Python Kompute examples.
|
||||
|
||||
## Running the Pi headless
|
||||
|
||||
|
|
@ -22,7 +22,7 @@ sudo apt-get update
|
|||
sudo apt-get upgrade
|
||||
```
|
||||
|
||||
## Install dependencies for building mesa and running Vulkan Kompute
|
||||
## Install dependencies for building mesa and running Kompute
|
||||
|
||||
```
|
||||
sudo apt-get install \
|
||||
|
|
@ -43,7 +43,7 @@ git clone --depth 1 https://gitlab.freedesktop.org/mesa/mesa.git
|
|||
|
||||
## Build mesa
|
||||
|
||||
Use meson and ninja to build mesa using the Broadcom Vulkan drivers. For information on the Gallium drivers please see [V3D — The Mesa 3D Graphics Library latest documentation](https://docs.mesa3d.org/drivers/v3d.html).
|
||||
Use meson and ninja to build mesa using the Broadcom Vulkan SDK drivers. For information on the Gallium drivers please see [V3D — The Mesa 3D Graphics Library latest documentation](https://docs.mesa3d.org/drivers/v3d.html).
|
||||
|
||||
```
|
||||
meson --libdir lib \
|
||||
|
|
@ -59,7 +59,7 @@ ninja -C build
|
|||
sudo ninja -C build install
|
||||
```
|
||||
|
||||
## Configure preferred Vulkan driver
|
||||
## Configure preferred Vulkan SDK driver
|
||||
|
||||
Export the path for the Broadcom drivers, this command will need to be run for every new terminal session.
|
||||
|
||||
|
|
@ -83,7 +83,7 @@ sudo chmod ugo+rw /dev/dri/renderD128
|
|||
sudo chown root:video /dev/dri/renderD128
|
||||
```
|
||||
|
||||
## Confirm correct Vulkan operation
|
||||
## Confirm correct Vulkan SDK operation
|
||||
|
||||
To confirm that mesa was configured and built correctly run the following command.
|
||||
|
||||
|
|
@ -91,9 +91,9 @@ To confirm that mesa was configured and built correctly run the following comman
|
|||
vulkaninfo
|
||||
```
|
||||
|
||||
## Clone Vulkan Kompute
|
||||
## Clone Kompute
|
||||
|
||||
Clone Vulkan Kompute for access to the latest Python tests.
|
||||
Clone Kompute for access to the latest Python tests.
|
||||
|
||||
```
|
||||
git clone https://github.com/EthicalML/vulkan-kompute.git
|
||||
|
|
@ -114,13 +114,13 @@ pip install git+git://github.com/EthicalML/vulkan-kompute.git
|
|||
|
||||
## Run the available tests
|
||||
|
||||
Use the following command to run the python tests for Vulkan Kompute.
|
||||
Use the following command to run the python tests for Kompute.
|
||||
|
||||
```
|
||||
pytest
|
||||
```
|
||||
|
||||
If the tests pass then congratulations! You are now able to make full use of the Pi 4 Broadcom GPU for running parallel computing. If however, there are any issues with the tests they can be run in debug mode to see the Vulkan logs.
|
||||
If the tests pass then congratulations! You are now able to make full use of the Pi 4 Broadcom GPU for running parallel computing. If however, there are any issues with the tests they can be run in debug mode to see the logs.
|
||||
|
||||
```
|
||||
pytest --log-cli-level debug
|
||||
|
|
@ -130,10 +130,10 @@ Please share any issues with the maintainers and they will be more than happy to
|
|||
|
||||
## Closing remarks
|
||||
|
||||
To avoid the need to export `VK_ICD_FILENAMES` every time you login, it is possible to symlink the json file into the default directory. The Vulkan loader looks in the `/etc/vulkan/icd.d/` directory for `.json` files.
|
||||
To avoid the need to export `VK_ICD_FILENAMES` every time you login, it is possible to symlink the json file into the default directory. The Vulkan SDK loader looks in the `/etc/vulkan/icd.d/` directory for `.json` files.
|
||||
|
||||
```
|
||||
sudo ln -s /mesa-install/share/vulkan/icd.d/broadcom_icd.armv7l.json /etc/vulkan/icd.d/broadcom_icd.armv7l.json
|
||||
```
|
||||
|
||||
As a word of warning, configuring the icd filenames in this way will stop certain tests being skipped. At the time of writing this will mean that some tests fail when running on the Pi.
|
||||
As a word of warning, configuring the icd filenames in this way will stop certain tests being skipped. At the time of writing this will mean that some tests fail when running on the Pi.
|
||||
|
|
|
|||
|
|
@ -1,3 +1,3 @@
|
|||
# Python Package for Vulkan Kompute
|
||||
# Python Package for Kompute
|
||||
|
||||
|
||||
|
|
|
|||
|
|
@ -191,7 +191,7 @@ load from physicalDevice)doc";
|
|||
|
||||
static const char *__doc_kp_Manager_Manager_3 =
|
||||
R"doc(Manager constructor which allows your own vulkan application to
|
||||
integrate with the vulkan kompute use.
|
||||
integrate with the kompute use.
|
||||
|
||||
@param instance Vulkan compute instance to base this application
|
||||
@param physicalDevice Vulkan physical device to use for application
|
||||
|
|
|
|||
2
setup.py
2
setup.py
|
|
@ -72,7 +72,7 @@ setup(
|
|||
name='kp',
|
||||
version='0.7.0',
|
||||
author='Alejandro Saucedo',
|
||||
description='Vulkan Kompute: Blazing fast, mobile-enabled, asynchronous, and optimized for advanced GPU processing usecases.',
|
||||
description='Kompute: Blazing fast, mobile-enabled, asynchronous, and optimized for advanced GPU processing usecases.',
|
||||
long_description=long_description,
|
||||
long_description_content_type='text/markdown',
|
||||
ext_modules=[CMakeExtension('kp')],
|
||||
|
|
|
|||
|
|
@ -2046,7 +2046,7 @@ class Manager
|
|||
|
||||
/**
|
||||
* Manager constructor which allows your own vulkan application to integrate
|
||||
* with the vulkan kompute use.
|
||||
* with the kompute use.
|
||||
*
|
||||
* @param instance Vulkan compute instance to base this application
|
||||
* @param physicalDevice Vulkan physical device to use for application
|
||||
|
|
|
|||
|
|
@ -145,8 +145,8 @@ Manager::createInstance()
|
|||
this->mFreeInstance = true;
|
||||
|
||||
vk::ApplicationInfo applicationInfo;
|
||||
applicationInfo.pApplicationName = "Vulkan Kompute";
|
||||
applicationInfo.pEngineName = "VulkanKompute";
|
||||
applicationInfo.pApplicationName = "Kompute";
|
||||
applicationInfo.pEngineName = "Kompute";
|
||||
applicationInfo.apiVersion = KOMPUTE_VK_API_VERSION;
|
||||
applicationInfo.engineVersion = KOMPUTE_VK_API_VERSION;
|
||||
applicationInfo.applicationVersion = KOMPUTE_VK_API_VERSION;
|
||||
|
|
|
|||
|
|
@ -39,7 +39,7 @@ class Manager
|
|||
|
||||
/**
|
||||
* Manager constructor which allows your own vulkan application to integrate
|
||||
* with the vulkan kompute use.
|
||||
* with the kompute use.
|
||||
*
|
||||
* @param instance Vulkan compute instance to base this application
|
||||
* @param physicalDevice Vulkan physical device to use for application
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue