NVIDIA AD102 Graphics Processing Unit (GPU)

Specs based on NVIDIA L40S

Most Powerful Variant

Cores : TMUs : ROPs

18176 : 568 : 192

Typical Memory Size

Typical Memory Bus

AD102 GPU

Manufacturer

NVIDIA

Architecture

Ada

Fabrication Process

4 nm

Die Size

608 mm²

Transistors Count

76.3B

Transistors Density

125.5M ^TRAN/_mm²

Tensor Cores

568

RT Cores

142

CUDAs

18176

SM

142

TMUs

568

ROPs

192

Boost Clock

2520 MHz

Memory Size

49152 MB

Memory Type

GDDR6

Memory Bus Width

384-bit

Memory Bandwidth

0.0 GB/s

Pixel Fillrate

483.8 GPixels/s

Texture Fillrate

1.4 TTexel/s

SPFP Performance

91.6 TFLOPS

Performance per W

261.7 GFLOPS/W

Performance per mm²

150.7 GFLOPS/mm²

Interface

PCI-Express 4.0 x16

Typical Thermal Design Power

350 W

SLI Support

2-way

DirectX

12.2

Vulkan

1.2

OpenGL

4.6

OpenCL

3.0

Pictures

Front

AD102 GPU Cards7

	Model	GPU Model	Unified Cores	Boost Clock	Memory Clock	Memory Config.
	NVIDIA TITAN RTX ADA	AD102	TBC	TBC	TBC	TBC
	NVIDIA RTX 6000 ADA	AD102	18176	2506 MHz	20 Gbps	48 GB G6 384b
	NVIDIA L40	AD102-895/895A	18176	2490 MHz	-	48 GB G6 384b
	NVIDIA L40S	AD102-895/895A	18176	2520 MHz	-	48 GB G6 384b
	NVIDIA GeForce RTX 4090 Ti	AD102	TBC	TBC	TBC	TBC
	NVIDIA GeForce RTX 4090	AD102-300	16384	2520 MHz	21 Gbps	24 GB G6X 384b
	NVIDIA RTX 5000 ADA	AD102	12800	2598 MHz	18 Gbps	32 GB G6 256b