AMD Boosting AI/LLM Performance For Radeon iGPUs As Much As 18~23% With Linux 7.4

Written by Michael Larabel in Display Drivers on 29 September 2026 at 02:57 PM EDT. Page 2 of 4. 23 Comments.
Lemonade Strix Halo

First up for testing was the AMD Ryzen AI Max+ 395 "Strix Halo" flagship setup. Benchmarks were done off the Framework Desktop with the AMD Ryzen AI Max+ 395 paired with 64GB of LPDDR5-8000 memory.

Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-debug, Model: Qwen3.5-4B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-short, Model: Qwen3.5-4B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: AMD ROCm, Scenario: code-short, Model: Qwen3-14B-GGUF. PerfOpt Disabled was the fastest.
Lemonade benchmark with settings of Backend: AMD ROCm, Scenario: code-short, Model: Qwen3-14B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-debug, Model: MiniCPM4-8B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-debug, Model: MiniCPM4-8B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-explain, Model: Qwen3-14B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-explain, Model: Qwen3-14B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-short, Model: MiniCPM4-8B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-short, Model: MiniCPM4-8B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-explain, Model: Qwen3.5-4B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-explain, Model: Qwen3.5-4B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: chat-long-output, Model: Qwen3.5-4B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: chat-long-output, Model: Qwen3.5-4B-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-short, Model: Qwen3-Coder-Next-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-short, Model: Qwen3-Coder-Next-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-explain, Model: Qwen3-Coder-Next-GGUF. PerfOpt was the fastest.
Lemonade benchmark with settings of Backend: Vulkan, Scenario: code-explain, Model: Qwen3-Coder-Next-GGUF. PerfOpt Disabled was the fastest.

On the AMD Ryzen AI Max+ desktop, Lemonade was typically seeing 2~4% better performance with PerfOpt as will be enabled with Linux 7.4... Not bad, but let's move on down the stack to other AMD iGPUs in laptops.

Related Articles
The Performance Benefits To Upgrading Mesa With Intel Arc B390 On Ubuntu Linux
Nouveau vs. NVIDIA R610 On CachyOS With The GeForce RTX 5090 Laptop GPU
Intel Arc Pro B70 Showing Off Some Performance Wins With Linux 7.1
Linux 7.1 Helping Intel Arc Battlemage Graphics Achieve Better Performance
ROCm 7.0.0 vs. ROCm 7.2.3 Performance On The AMD Radeon AI PRO R9700
Nouveau vs. NVIDIA R595 Linux Driver For Workstation Graphics Performance