QuData

Rent MI300X by the Hour by QuData Ai

Explore the specifications and benchmarks of the MI300X. Compare hardware configurations and performance metrics to optimize your cutting-edge AI and HPC workloads with AMD's absolutely flagship GPU architecture.

GPU Description

Explore the specifications and benchmarks of the MI300X. Compare hardware configurations and performance metrics to optimize your cutting-edge AI and HPC workloads with AMD's absolutely flagship GPU architecture.

GPU Characteristics

GPU typeMI300X
Connecting GPUOAM
GPU model nameAMD Instinct MI300X
Cooling systemLiquid Cooling
GPU memory192GB HBM3
Power Limit (W)750.0
GPU temperature (°C)95
GPU Clock Speed (MHz)2100
Memory Clock Speed (MHz)1300
Memory bandwidth5325 GB/s

GPU Configuration

Stream Processors19456
Matrix of the core304
Support for FP16/INT8Yes
ArchitectureCDNA 3
VirtualizationYes

Benchmarks

ffmpeg~6ms
Coremark (iterations per second)~225,000-295,000
Llama2 Inference (tokens per second)~585-765
Tensorflow Mixed Training~35.5-48.5
Stable Diffusion (images/min)~365-485
AI Training (TFLOPS FP64)~83.7

Providers with MI300X