Performance Features

NVIDIA Ampere Architecture

NVIDIA RTX A400 is an efficient, single slot professional solution for CAD, DCC, financial service industry (FSI) and visualization professionals in general looking to reach great performance in a compact form factor. Building upon the major SM enhancements from the Turing GPU, the NVIDIA Ampere architecture enhances ray tracing operations, tensor matrix operations, and concurrent executions of FP32 and INT32 operations.

CUDA Cores

The NVIDIA Ampere architecture-based CUDA cores bring up to 2.5X the single-precision floating point (FP32) throughput compared to the previous generation, providing significant performance improvements for graphics workflows such as 2D graphics, 3D model development, basic photo and video editing and compute for workloads such as data analysis and general productivity. RTX A400 enables two FP32 primary data paths, doubling the peak FP32 operations.

2nd Generation RT Cores

Incorporating 2nd generation ray tracing engines, NVIDIA Ampere architecture-based GPUs provide incredible ray traced rendering performance. For the first time, NVIDIA is introducing RT Cores into a low profile form factor GPU. A single RTX A400 board can render complex professional models with physically accurate shadows, reflections, and refractions to empower users with instant insight. Working in concert with applications leveraging APIs such as NVIDIA OptiX, Microsoft DXR, and Vulkan ray tracing, systems based on RTX A400 will power truly interactive design workflows to provide immediate feedback for unprecedented levels of productivity. RTX A400 is up to 4X faster in rendering performance compared to the previous generation. This technology also speeds up the rendering of ray-traced motion blur for faster results with greater visual accuracy.

3rd Generation Tensor Cores

Purpose-built for design and productivity applications, RTX A400 includes enhanced Tensor Cores as part of the Ampere GPU architecture that accelerate more datatypes that deliver up to 2.5X the generative AI performance compared to the previous generation. 3rd-gen Tensor Cores will accelerate the TF32 and BFloat16 precision modes. Independent floating-point and integer data paths allow more efficient execution of workloads using a mix of computation and addressing calculations.

NVIDIA RTX A400

• NVIDIA Ampere architecture
• 768 NVIDIA® CUDA® Cores
• 24 NVIDIA® Tensor Cores
• 6 NVIDIA® RT Cores
• 4GB GDDR6 Memory
• 96GB/s Memory Bandwidth
• Max. Power Consumption: 50W
• Graphics Bus: PCI-E 4.0 x8
• Thermal Solution: Active Fan
• Display Connectors: mDP 1.4a (4)

Specifications

Specifications

GPU Architecture NVIDIA Ampere Architecture

CUDA Cores 768

NVIDIA Tensor Cores 24

NVIDIA RT Cores 6

Single-Precision Performance1 2.7 TFLOPS

RT Core Performance1 5.3 TFLOPS

Tensor Performance1 21.7 TFLOPS2

GPU Memory 4 GB GDDR6

Memory Interface 64-bit

Memory Bandwidth 96 GB/s

Max Power Consumption 50W

Graphics Bus PCI Express 4.0 x83

Display Connectors mDP 1.4a (4)

Form Factor 2.7” H x 6.4” L Single Slot

Product Weight 140 g (Low Profile Bracket)

146 g (ATX Bracket)

Thermal Solution Active Fan

NVENC | NVDEC 1x | 1x (+AV1 decode)