While the Ryzen 5 5600G has taken the gaming world by storm, replacing its predecessor, the Ryzen 5 4600G, in terms of CPU performance, a recent discovery has given the older chip a new lease on life. With a clever trick, users can now turn the budget-friendly Zen 2 APU into a 16GB graphics card, enabling it to run AI applications effectively on Linux.
In an industry where AI-focused graphics cards have become increasingly sought after, not everyone has the means to acquire the likes of the expensive Nvidia H100 (Hopper) or A100 (Ampere). These high-end options may be in high demand, and some enthusiasts may find it nearly impossible to get their hands on them. However, one innovative Redditor has demonstrated that you don’t need these costly graphics cards to delve into AI experimentation.
The Ryzen 5 4600G, released in 2020, features a hexa-core, 12-thread APU with Zen 2 cores that operate at a base and boost clock speed of 3.7 GHz and 4.2 GHz, respectively. In addition, this 65W chip is equipped with a Radeon Vega iGPU with seven compute units clocked up to 1.9 GHz. It’s worth noting that APUs don’t have dedicated memory and instead share system memory. In this particular case, the Redditor had 32GB of DDR4 and allocated 16GB to the Ryzen 5 4600G, the maximum amount typically allowed for an iGPU.
What’s truly remarkable about this trick is that it transforms the Ryzen 5 4600G into a “graphics card” with 16GB of memory. In fact, it boasts more memory than some of Nvidia’s latest GeForce RTX 40-series SKUs, which are restricted to just 12GB. Although the APU cannot match the performance of high-end graphics cards, it offers enough memory for non-serious AI tasks and ensures that users won’t encounter any memory limitations.
Initially, AMD’s Radeon Open Compute platform (ROCm) did not support Ryzen APUs officially. However, third-party companies like BruhnBruhn Holding have developed experimental ROCm packages that work effectively with APUs. This advancement means that APUs can now be utilized with popular AI software frameworks like PyTorch and TensorFlow, potentially unlocking a wide range of possibilities for these chips.
The Redditor shared a video showcasing the Ryzen 5 4600G’s capabilities with various AI applications, such as Stable Diffusion, FastChat, MiniGPT-4, Alpaca-LoRA, Whisper, LLM, and LLaMA. Although the video only demonstrated the capabilities of Stable Diffusion, an AI image generator, the Redditor promises to release a comprehensive video detailing the setup process in the future.
Ultimately, this experiment highlights the potential of budget CPUs like the Ryzen 5 4600G and Ryzen 5 5600G for AI enthusiasts interested in exploring new frontiers. However, for individuals without these specific processors, investing $500 in an APU build might not be the most sensible choice, particularly when there are discrete graphics cards available that offer superior performance. With AMD’s Radeon 16GB graphics cards starting at $499 and Nvidia’s recent launch of the GeForce RTX 4060 Ti 16GB, users have alternative options that may better suit their needs.
Frequently Asked Questions (FAQ)
Can I transform the Ryzen 5 4600G into a 16GB graphics card?
Yes, with a clever trick, you can allocate system memory to the Ryzen 5 4600G, essentially transforming it into a 16GB “graphics card.” This allows it to handle AI applications effectively on a Linux system.
How does the Ryzen 5 4600G compare to its successor, the Ryzen 5 5600G, for gaming?
The newer Ryzen 5 5600G has replaced the Ryzen 5 4600G as one of the best CPUs for gaming. It offers improved performance and is often preferred by gamers seeking the best possible experience.
Can I use AMD’s Radeon Open Compute platform (ROCm) with Ryzen APUs?
While AMD’s ROCm does not officially support Ryzen APUs, third-party companies such as BruhnBruhn Holding offer experimental packages of ROCm that are compatible with APUs. This allows Ryzen APUs to work effectively with popular AI software frameworks like PyTorch and TensorFlow.
Are there alternative options available for AI enthusiasts without a Ryzen 5 4600G or Ryzen 5 5600G?
Absolutely. If you don’t already own a Ryzen 5 4600G or Ryzen 5 5600G, investing $500 in an APU build may not be the best choice. Instead, you can consider opting for discrete graphics cards such as AMD’s Radeon 16GB graphics cards, which start at $499, or Nvidia’s recently launched GeForce RTX 4060 Ti 16GB, which offers similar performance.