GeekRadar / The comparison desk

COMPARE.

GPU

NVIDIA

NVIDIA Jetson AGX Xavier 32 GB

Graphics processor

Video memory: 32 GB^*
Memory bus: 256 bit^*
Rated memory bandwidth: 136.50 GB/s^*

No matched benchmark results are available for this entry yet.

Processing & platform

Architecture: Volta^*
Process node: 12 nm^*
GPU cores: 512^*
Graphics interface: PCIe 4.0 x4^*
Reported chip die area: 350 mm²^*
Transistors: 9,000 million^*
Graphics processor: GV10B^*
Texture mapping units: 32^*
Raster output units: 16^*
Streaming multiprocessors: 8^*
Tensor cores: 64^*
Chip foundry: TSMC^*
Card or module format: IGP^*
Declared GPU TDP: 30 W^*
DirectX / feature level: 12 (12_1)^*
OpenGL: 4.6^*
OpenCL: 1.2^*
Vulkan: 1.2^*
CUDA compute capability: 7.2^*
Shader model: 6.0^*
GPU L2 cache: 512 KB^*

Memory & storage

Memory options: 32^*
Video memory: 32 GB^*
Video memory type: LPDDR4X^*
Memory bus: 256 bit^*
Rated memory bandwidth: 136.50 GB/s^*

Power specifications

Thermal design power: 30 W^*

System configuration

Detailed specifications

Specifications 1 · 28 fields

proc transistors: 9,000 million^*
proc die size: 350^* mm²
cores: 512^*
tensor cores: 64^*
sm count: 8^*
base clock: 854^*
boost clock: 1,377^*
memory clock: 2,133^*
ram: 32^*
name: NVIDIA Jetson AGX Xavier 32 GB^*
ram type: LPDDR4X^*
bus width: 256^*
bus interface: PCIe 4.0 x4^*
ram bandwidth: 136.5^* GB/s
fp16 performance: 2.82^*
fp32 performance: 1.41^*
fp64 performance: 0.705^*
tdp: 30^* W
slot width: IGP^*
api cuda: 7.2^*
is retail board: No^*
gpi value: 2.3^*
vendor: nvd^*
manufacturer: NVIDIA^*
arch name: Volta^*
card release date: 2018-10-13^*
proc foundry: TSMC^*
proc process size: 12^* nm

Specifications 2 · 37 fields

CUDA compute capability

7.2^*

DirectX / feature level

12 (12_1)^*

OpenCL

1.2^*

OpenGL

4.6^*

Shader model

6.0^*

Vulkan

1.2^*

Architecture

Volta^*

Graphics interface

PCIe 4.0 x4^*

Memory bus

256^* bit

Reported shader processors

512^*

Reported theoretical FP16 peak

2.82^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Reported theoretical FP32 peak

1.41^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Reported theoretical FP64 peak

0.705^* TFLOPS

Scope and conditions

Arithmetic density: Not stated^*

Declared L1 cache configuration

128 KB (per SM)^*

Declared GPU L2 cache

512 KB^*

Reported display outputs

Portable Device Dependent^*

Theoretical pixel fill rate

22.03 GPixel/s^*

Reported chip die area

350 mm²^*

Chip foundry

TSMC^*

Process node declaration

12 nm^*

Transistors

9,000 million^*

Graphics processor

GV10B^*

Reported video memory

32^* GB

Rated memory bandwidth

136.5 GB/s^*

Video memory type

LPDDR4X^*

Raster output units

16^*

Reported shading units

512^*

Card or module format

IGP^*

Streaming multiprocessors

8^*

2:4 tensor sparsity

No^*

Declared GPU TDP

30 W^*

Tensor cores

64^*

Theoretical texture fill rate

44.06 GTexel/s^*

Texture mapping units

32^*

GPU cores

512^*

Video memory

32^* GB

Video memory type

LPDDR4X^*

* Provisional data