Instinct MI300X 192GB

AMD · 192GB · 2 reports

This page is thin (2 of 3 reports needed for indexing). Help fill it in.

Latest Most reported Fastest t/s

Gemma 4 25.2B (3.8B active) 26B Instruct (MoE)

Instinct MI300X 192GB · vLLM

throughput:: 60.0 t/s gen
quant:: FP16 (safetensors)

text-generation

Gemma 4 26B MoE on Instinct MI300X via vLLM + ROCm. Full FP16. MoE active params make it very fast. Source: gemma4-ai.com AMD GPU guide

Gemma 4 30.7B Instruct

Instinct MI300X 192GB · vLLM

throughput:: 35.0 t/s gen
quant:: FP16 (safetensors)

text-generation

Gemma 4 31B on Instinct MI300X via vLLM + ROCm. Full FP16. Datacenter grade. Source: gemma4-ai.com AMD GPU guide