RunLocalAI
amd
GPU
288GB VRAM
workstation

AMD Instinct MI355X

Latest CDNA 4. 288GB HBM3e — currently the highest VRAM per chip on the market.

Released 2025

Overview

Latest CDNA 4. 288GB HBM3e — currently the highest VRAM per chip on the market.

Specs

VRAM288 GB
Power draw1000 W
Released2025
MSRP$25000
Backends
ROCm

Models that fit

Open-weight models small enough to run on AMD Instinct MI355X with usable context.

Llama 3.1 8B Instruct
8B · llama
Llama 4 Scout
109B · llama
Qwen 3 30B-A3B
30B · qwen
Qwen 2.5 Coder 32B Instruct
32B · qwen
Llama 3.3 70B Instruct
70B · llama
Qwen 3 32B
32B · qwen
Gemma 4 31B Dense
31B · gemma
Qwen 3 8B
8B · qwen

Frequently asked

What models can AMD Instinct MI355X run?

With 288GB VRAM, the AMD Instinct MI355X runs 70B models in 4-bit quantization, plus everything smaller. See the model list below for tested combinations.

Does AMD Instinct MI355X support CUDA?

No — AMD Instinct MI355X is an AMD card. Use ROCm (Linux) or the Vulkan backend in llama.cpp instead. CUDA-only tools won't work.

Reviewed by RunLocalAI Editorial. See our editorial policy for how we research and verify hardware specifications.