AMD's Ryzen AI Boosts Performance by 27% in AI Tasks

AMD has shown incredible performance in AI tasks with its new Strix Point APUs. These processors are giving Intel's Lunar Lake chips a run for their money, especially in large language model (LLM) applications. For instance, the Ryzen AI 9 HX 375 can handle LLM tasks up to 27% faster than Intel's Core Ultra 7 258V in consumer applications. This isn't just about speed; latency is also reduced by a whopping 3. 5 times! The Ryzen AI 300 series isn’t just swift in computations; it also packs powerful integrated graphics. These iGPUs can offload work and boost performance using the Vulkan API. They’re based on AMD's RDNA 3. 5 architecture and can improve performance for LLM tasks by up to 31%.

To ensure a fair comparison, AMD tested these processors in Intel's AI Playground. The results? The Ryzen AI 9 HX 375 was up to 13% faster on some tasks compared to the Intel chip. With features like Variable Graphics Memory (VGM), which shifts work to the iGPU, AMD ensures better performance and power efficiency. AMD is making AI accessible to everyone with tools like LM Studio. This user-friendly software is built on the llama. cpp framework and doesn't require deep technical knowledge. It's all about making AI easier for everyone.

AMD's Ryzen AI Boosts Performance by 27% in AI Tasks – Outpacing Intel's Lunar Lake

Actions