QuData

Rent A10 by the Hour by QuData Ai

Explore the specifications and benchmarks of A10. Compare hardware configurations and performance metrics for cloud computing and virtualization with the Ampere architecture for data centers.

GPU Description

Explore the specifications and benchmarks of A10. Compare hardware configurations and performance metrics for cloud computing and virtualization with the Ampere architecture for data centers.

GPU Characteristics

GPU typeA10
Connecting GPUPCIe
GPU model nameNVIDIA A10
Driver version535.183.01
GPU memory24GB GDDR6
Power Limit (W)150.0
GPU temperature (°C)65
GPU Clock Speed (MHz)1695
Memory Clock Speed (MHz)1562
Memory bandwidth600 GB/s

GPU Configuration

CUDA cores9216
Tensor core288 (3rd generation)
Support for FP16/INT8Yes
ArchitectureAmpere
RT core72 (2nd generation)

Benchmarks

ffmpeg~8ms
Coremark (iterations per second)~58000-68000
Llama2 Inference (tokens per second)~45-55
Tensorflow Mixed Training~4.2-5.1
Stable Diffusion (images/min)~28-35
Mixed Precision Training~5.8-7.2x

Providers with A10