Skip to main content
HardwareAI InfrastructurevMI325X

AMD Instinct MI325X

by AMD · paid · Last verified 2026-03-17

Enhanced refresh of the MI300X with 256GB HBM3e memory and higher memory bandwidth. Targets hyperscalers and enterprises seeking maximum memory capacity for inference of the largest open models like Llama and Mixtral.

https://www.amd.com/en/products/accelerators/instinct/mi300/mi325x.html
C
CBelow Average
Adoption: CQuality: A+Freshness: ACitations: C+Engagement: F

Specifications

License
Proprietary
Pricing
paid
Capabilities
ai-training, inference, fp8-compute, high-capacity-memory
Integrations
rocm, pytorch, tensorflow, vllm
Use Cases
large-model-inference, llm-training, hpc
API Available
No
Tags
gpu, data-center, training, inference, amd, cdna3
Added
2026-03-17
Completeness
100%

Index Score

49.5
Adoption
42
Quality
91
Freshness
88
Citations
58
Engagement
0

Put AI to work for your business

Deploy this hardware alongside autonomous AaaS agents that handle tasks end-to-end — no babysitting required.

Explore the full AI ecosystem on Agents as a Service