AMD’s Radeon RX 7900 XTX Surpasses NVIDIA’s GeForce RTX 4090 in DeepSeek’s AI Inference Tests; Learn How to Implement R1 on Your Local AMD Devices

The performance of AMD’s Radeon RX 7900 XTX with DeepSeek’s R1 AI model has truly impressed the gaming industry, surpassing the benchmarks set by NVIDIA’s GeForce RTX 4090 in inference tests.

AMD Rapidly Brings Support for DeepSeek’s R1 AI Model, Offering Exceptional Performance

DeepSeek’s latest AI model has quickly made waves, and while the tech community buzzes about the resources needed to train such models, it turns out an average gamer can achieve outstanding performance using AMD’s “RDNA 3” Radeon RX 7900 XTX GPU. AMD has shared inference benchmark results for DeepSeek’s R1, showcasing how their leading RX 7000 series GPU outperforms NVIDIA’s equivalent in multiple scenarios.

For tech enthusiasts who run AI workloads on their consumer GPUs, this is great news. Such GPUs not only provide excellent value but also ensure users can maintain their privacy—a significant advantage, particularly with the privacy concerns surrounding DeepSeek’s AI models. Luckily, AMD has released a detailed guide on deploying DeepSeek R1 models on their GPUs. Here’s how you can do it:

Step 1: Ensure you’re using the 25.1.1 Optional or newer Adrenalin driver.

Step 2: Grab LM Studio 0.3.8 or later from lmstudio.ai/ryzenai.

Step 3: Install LM Studio and bypass the onboarding steps.

Step 4: Navigate to the discover tab.

Step 5: Choose your DeepSeek R1 Distill. Smaller models like Qwen 1.5B are lightning-fast and serve as a great starting point, while larger models offer enhanced reasoning capabilities.

Step 6: Ensure “Q4 K M” quantization is selected on the right, then click “Download”.

Step 7: After downloading, return to the chat tab, select the DeepSeek R1 distillation from the dropdown, and enable “manually select parameters”.

Step 8: Max out the GPU offload layers by sliding the bar to the end.

Step 9: Initiate the model load.

Step 10: Engage with a reasoning model entirely powered by your local AMD hardware!

Should these instructions seem unclear, AMD has thoughtfully provided a YouTube tutorial that breaks down each step, making it easy to run DeepSeek’s LLMs on your AMD devices—securing your data and ensuring it’s not misused. As we look to future releases from NVIDIA and AMD, we anticipate even greater inferencing abilities, driven by the sophisticated AI engines integrated into these forthcoming GPUs.