Specifications
Architecture
CDNA 4
Memory
144GB HBM3E
Memory Bandwidth
4 TB/s peak · 3.6 TB/s delivered
Power
Up to 600 W
Form Factor
PCIe add-in card
Target
AI inference / decode
Compute
MI350 Series
Overview
The AMD Instinct MI350P brings 144GB of HBM3E on a PCIe add-in card, letting standard 2U servers run large-model inference and decode without OAM infrastructure. QS Compute supplies MI350 Series accelerators for AI data center deployments.
Request a Quote — AMD Instinct MI350P — 144GB HBM3E PCIe AI Accelerator
QS Compute — Global B2B AI computing hardware supply, 15-day sample lead time.
Get Your Quote →