RTX 4080

Ada Lovelace2022320W TDPBF16FP8
NVIDIA Product Page

Specifications

CUDA Cores9,728
Tensor Cores304
AI TOPS780
FP32 TFLOPS49.0
Memory16 GB GDDR6X
Memory Bandwidth717 GB/s
Memory Interface256-bit
TDP320W

Pricing

Proshop (cheapest)-
Vast.ai$0.23/hr
Google Cloud-
Modal-

Efficiency

Power Efficiency
CUDA/Watt
30.40
Tensor/Watt
0.95
Bandwidth/Watt
2.24GB/s/W
Rental Value (CUDA/$/hr)
Vast.ai
42,295.65