tostc

Date of the guide : September, 2025

Introduction

In this post, I will provide the solution that worked on my system on how to install Radeon Open Compute (ROCm) on Arch (linux-6.6.7.arch1-1) for RX 6900 XT (Should work on other 6000 series). ROCm is an open-source software platform that allows GPU-accelerated computation. This tool is a prerequist to use GPU Acceleration on TensorFlow or PyTorch. In this guide I will use Paru as my AUR package helper, feel free to use any other (https://wiki.archlinux.org/title/AUR_helpers). I will assume you have a working operating system and know what you do with it (Otherwise Arch will be painfull for you).

Jellyfin API Authorization

To start using the Jellyfin API, authorization is probably the first thing you'll need to do. Jellyfin's authorization options can be a bit confusing because there are a lot of deprecated options.

Generally there are three ways to authenticate: no authorization, user authorization with an access token or authorization with an API key. The first way is easy, just do nothing. But most often you'll need to use either the access token or API key.

Sending authorization values

There are multiple methods for transmitting authorization values, however, some are outdated and scheduled to be removed. It's recommend to use the Authorization header. If header auth isn't an option, the token may be sent through the ApiKey query parameter. Sending secure data in a query parameter is unsafe as the changes of it leaking (via logs, copy-paste actions or by other means) are high. Only use this method as a last resort.

GPU Optimization for GameDev

Graphics Pipeline / GPU Architecture Overview

2011 - A trip through the Graphics Pipeline 2011
2013 - Performance Optimization Guidelines and the GPU Architecture behind them
2015 - Life of a triangle - NVIDIA's logical pipeline
2015 - Render Hell 2.0
2016 - How bad are small triangles on GPU and why?
2017 - GPU Performance for Game Artists
2019 - Understanding the anatomy of GPUs using Pokémon

	tool
	extends SceneTree

	const PLATFORM_ANDROID = "Android"
	const PLATFORM_HTML5 = "HTML5"
	const PLATFORM_IOS = "iOS"
	const PLATFORM_LINUX = "Linux/X11"
	const PLATFORM_MAC = "Mac OSX"
	const PLATFORM_UWP = "UWP"
	const PLATFORM_WINDOWS = "Windows Desktop"

	// ln.c
	//
	// simple, fast, accurate natural log approximation
	// when without <math.h>

	// featuring * floating point bit level hacking,
	// * x=m*2^p => ln(x)=ln(m)+ln(2)p,
	// * Remez algorithm

	// by Lingdong Huang, 2020. Public domain.

	# This is a basic workflow to help you get started with Actions
	# workflow - цепочка действий
	# Имя процесса Билдится на всех типах 📦 🐍
	name: CMake Build Matrix

	# Controls when the action will run. Triggers the workflow on push
	on:
	push:
	pull_request:
	release:

	#include <assert.h>
	#include <stdio.h>
	#include <stdlib.h>

	#include <X11/Xlib.h> /* of course */
	#include <X11/Xutil.h> /* duuuh */
	#include <sys/ipc.h>
	#include <sys/shm.h>
	#include <X11/extensions/XShm.h> /* obviously */

	import discord
	import asyncio
	import speech_recognition as sr
	from threading import Thread

	# bot token and wit.ai api key
	TOKEN = ""
	WIT_AI_KEY = ""

	# we need a sink for the listen function, so we just define our own

	#ifndef PLATFORM
	#define PLATFORM

	#define GCC_COMPILER 1
	#define MVSC_COMPILER 2
	#define ICC_COMPILER 3
	#define BORLAND_COMPILER 4

	#define LINUX 1
	#define WINDOWS 2

	#include <iostream>
	#include <SDL.h>

	using namespace std;

	int main() {
	if(SDL_Init(SDL_INIT_VIDEO) != 0) {
	std::cout << "SDL_Init Error: " << SDL_GetError() << std::endl;

	return 1;