GeekRadar / The comparison desk

COMPARE.

GPU

NVIDIA

NVIDIA A10 PCIe

Graphics processor

Video memory: 24 GB^*
Memory bus: 384 bit^*
Rated memory bandwidth: 600.20 GB/s^*

No matched benchmark results are available for this entry yet.

Processing & platform

Architecture: Ampere^*
Process node: 8 nm^*
GPU cores: 9,216^*
Reported GPU frequency: 885 MHz^*
Graphics interface: PCIe 4.0 x16^*
Reported chip die area: 628 mm²^*
Chip/package TDP: 150 W^*
Transistors: 28,300 million^*
GPU L2 cache: 6 MB^*
Graphics processor: GA102^*
Texture mapping units: 288^*
Raster output units: 96^*
Streaming multiprocessors: 72^*
Tensor cores: 288^*
Ray tracing cores: 72^*
Chip foundry: Samsung^*
Card or module format: Single-slot^*
Declared GPU TDP: 150 W^*
Suggested system PSU: 450 W^*
External power connectors: 1x 8-pin^*
DirectX / feature level: 12 Ultimate (12_2)^*
OpenGL: 4.6^*
OpenCL: 3.0^*
Vulkan: 1.4^*
CUDA compute capability: 8.6^*
Shader model: 6.9^*

Memory & storage

Memory options: 24^*
Video memory: 24 GB^*
Video memory type: GDDR6^*
Memory bus: 384 bit^*
Rated memory bandwidth: 600.20 GB/s^*

Power specifications

Thermal design power: 150 W^*

System configuration

Detailed specifications

Specifications 1 · 13 fields

Foundry: Samsung^*
Vendor: NVIDIA^*
FP16 GFLOPS: 31,240^* GFLOPS
FP32 GFLOPS: 31,240^* GFLOPS
FP64 GFLOPS: 976.3^* GFLOPS
Product: NVIDIA A10 PCIe^*
Type: GPU^*
Release Date: 2021-04-12^*
Process Size (nm): 8^* nm
TDP (W): 150^* W
Die Size (mm^2): 628^* mm²
Transistors (million): 28,300^*
Freq (MHz): 885^* MHz

Specifications 2 · 31 fields

proc transistors: 28,300 million^*
proc die size: 628^* mm²
cores: 9,216^*
tensor cores: 288^*
rt cores: 72^*
sm count: 72^*
base clock: 885^*
boost clock: 1,695^*
memory clock: 1,563^*
ram: 24^*
name: NVIDIA A10 PCIe^*
ram type: GDDR6^*
bus width: 384^*
bus interface: PCIe 4.0 x16^*
ram bandwidth: 600.2^* GB/s
fp16 performance: 31.24^*
fp32 performance: 31.24^*
fp64 performance: 0.9763^*
tdp: 150^* W
suggested psu: 450^* W
slot width: Single-slot^*
power connectors: 1x 8-pin^*
api cuda: 8.6^*
is retail board: No^*
gpi value: 16.93^*
vendor: nvd^*
manufacturer: NVIDIA^*
arch name: Ampere^*
card release date: 2021-04-12^*
proc foundry: Samsung^*
proc process size: 8^* nm

Specifications 3 · 40 fields

CUDA compute capability

8.6^*

DirectX / feature level

12 Ultimate (12_2)^*

OpenCL

3.0^*

OpenGL

4.6^*

Shader model

6.9^*

Vulkan

1.4^*

Architecture

Ampere^*

Graphics interface

PCIe 4.0 x16^*

Memory bus

384^* bit

Reported shader processors

9,216^*

Reported theoretical FP16 peak

31.24^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Reported theoretical FP32 peak

31.24^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Reported theoretical FP64 peak

0.9763^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Declared L1 cache configuration

128 KB (per SM)^*

Declared GPU L2 cache

6 MB^*

Reported display outputs

No outputs^*

Theoretical pixel fill rate

162.7 GPixel/s^*

External power connectors

1x 8-pin^*

Reported chip die area

628 mm²^*

Chip foundry

Samsung^*

Process node declaration

8 nm^*

Transistors

28,300 million^*

Graphics processor

GA102^*

Reported video memory

24^* GB

Rated memory bandwidth

600.2 GB/s^*

Video memory type

GDDR6^*

Raster output units

96^*

Ray tracing cores

72^*

Reported shading units

9,216^*

Card or module format

Single-slot^*

Streaming multiprocessors

72^*

Suggested system PSU

450 W^*

2:4 tensor sparsity

Yes^*

Declared GPU TDP

150 W^*

Tensor cores

288^*

Theoretical texture fill rate

488.2 GTexel/s^*

Texture mapping units

288^*

GPU cores

9,216^*

Video memory

24^* GB

Video memory type

GDDR6^*

* Provisional data