48 GB

GDDR6

2520 MHz

Boost Clock

275 W

TDP

5 nm

Process
Share on Social Media
Share L20 specifications with others

Technical Specifications

Core Specifications
Architecture: Ada Lovelace
Process Size: 5 nm
Base Clock: 1440 MHz
Boost Clock: 2520 MHz
Shader Units: 11776
SMs: 92
Tensor Cores: 368
RT Cores: 92
TMUs: 368
ROPs: 128
Memory & Performance
Memory Size: 48 GB
Memory Type: GDDR6
Memory Bandwidth: 864.0 GB/s
Bus Interface: PCIe 4.0 x16
TDP: 275 W
Suggested PSU: 600 W
Cache & Die Information
L1 Cache: 128 KB
L2 Cache: 96.0 MB
Die Size: 609 mm²
Transistors: 76.3 billion
Transistor Density: 125.3 million/mm²

API Support

DirectX
12.2
OpenGL
4.6
Vulkan
1.3
Cuda Compute
8.9 sm_89

Physical Specifications

Board Length: 267 mm
Board Width: 111 mm
Power Connectors: 1x 16-pin
Display Outputs: 4x DisplayPort 1.4a

Quick Facts

Release Date: 2023-11-16
Manufacturer: NVIDIA
Generation: Server Ada(Lxx)

Performance

Pixel Fill Rate 322.6 GPixel/s
Texture Fill Rate 927.4 GTexel/s
FP32 Performance 59.4 TFLOPS

CUDA Toolkit Compatibility

Cuda Compute
8.9 sm_89
Architecture

Ada Lovelace

Supported CUDA Toolkits
11.8 12.0 12.1 12.2 12.3 12.4 12.5 12.6 12.7 12.8

Ada Lovelace architecture with 3rd gen RT cores and AV1 encoding