Home NVIDIA PG506 242

NVIDIA PG506 242

This is a GPU manufactured with TSMC 7nm process, based on Nvidia Ampere architecture and released on Apr 2021. It features 54.2 billion transistors, 3584 CUDA cores and 24GB HBM2 memory, with 24MB L2 cache, theoretical performance of 10.32TFLOPS, with total power consumption of 165W.

Graphics Card

[Report Issues]
Release Date
Apr 2021
Generation
Tesla
Type
Desktop
Bus Interface
PCIe 4.0 x16

Clock Speeds

[Report Issues]
Base Clock
930 MHz
Boost Clock
1440 MHz
Memory Clock
1215 MHz

Memory

[Report Issues]
Memory Size
24GB
Memory Type
HBM2
Memory Bus
3072bit
Bandwidth
933.1GB/s

Render Config

[Report Issues]
SM Count
56
Shading Units
3584
TMUs
224
ROPs
96
Tensor Cores
224
L1 Cache
192 KB (per SM)
L2 Cache
24 MB

Theoretical Performance

[Report Issues]
Pixel Rate
138.2 GPixel/s
Texture Rate
322.6 GTexel/s
FP16 (half)
10.32 TFLOPS
FP32 (float)
10.32 TFLOPS
FP64 (double)
5.161 TFLOPS

Graphics Processor

[Report Issues]
GPU Name
GA100
Architecture
Ampere
Foundry
TSMC
Process Size
7 nm
Transistors
54.2 billion
Die Size
826 mm²

Board Design

[Report Issues]
TDP
165W
Suggested PSU
450 W
Outputs
No outputs
Power Connectors
8-pin EPS

Graphics Features

[Report Issues]
DirectX
N/A
OpenGL
N/A
OpenCL
3.0
Vulkan
N/A
CUDA
8.0
Shader Model
N/A

Rankings

[Report Issues]
FP32 (float)
10.44 TFLOPS
10.34 TFLOPS
NVIDIA A30 PCIe 24 GB HBM2e
10.32 TFLOPS
NVIDIA PG506 232 24 GB HBM2
10.32 TFLOPS
NVIDIA PG506 242 24 GB HBM2
10.32 TFLOPS
10.29 TFLOPS

Related Comparisons

Related News

© 2024 - TopCPU.net   Contact Us Privacy Policy