Model

QUADRO RTX4000 D6-8GB

GPU Architecture

Turing

CUDA Cores

2304

Tensor Cores

288

RT Cores

36

GPU Memory

8 GB GDDR6

RTX-OPS

431

Rays Cast

6 Giga Rays/ Sec

Single Precision (FP32) Performance

7.1 TFLOPS

Half Precision (FP16) Performance

14.2 TFLOPS

Integer Operation (INT8) Performance

28.5 TOPS

Deep Learning TeraFLOPS

57.0 Tensor TFLOPS

Memory Interface

256-bit

Memory Bandwidth

416 GB/s

Max Power Consumption

160 W