NVIDIA GeForce RTX 3090 vs NVIDIA H100
Side-by-side comparison of specifications, performance, and efficiency.
NVIDIA GeForce RTX 3090
OfficialArchitecture
Ampere
Year
2020
VRAM
24 GB
TDP
350 W
NVIDIA H100
OfficialArchitecture
Hopper
Year
2022
VRAM
80 GB
TDP
350 W
Specifications
NVIDIA GeForce RTX 3090 NVIDIA H100
Architecture
Ampere Hopper
Launch Year
2020 2022
Form Factor
PCIe PCIe
VRAM
24 GB80 GB
Memory Bandwidth
936 GB/s2.0 TB/s
TDP
350 W350 W
Process Node
8nm 4nm
| Spec | NVIDIA GeForce RTX 3090 | NVIDIA H100 |
|---|---|---|
| Architecture | Ampere | Hopper |
| Launch Year | 2020 | 2022 |
| Form Factor | PCIe | PCIe |
| VRAM | 24 GB | 80 GB |
| Memory Bandwidth | 936 GB/s | 2.0 TB/s |
| TDP | 350 W | 350 W |
| Process Node | 8nm | 4nm |
Performance (TFLOPS)
NVIDIA GeForce RTX 3090 NVIDIA H100
FP64
0.6 TFlops26.0 TFlops
FP32
35.6 TFlops51.0 TFlops
FP16
35.6 TFlops756.5 TFlops
BF16
35.6 TFlops756.5 TFlops
FP8
—1513.0 TFlops
INT8
71.2 TFlops1513.0 TFlops
| Precision | NVIDIA GeForce RTX 3090 | NVIDIA H100 |
|---|---|---|
| FP64 | 0.6 TFlops | 26.0 TFlops |
| FP32 | 35.6 TFlops | 51.0 TFlops |
| FP16 | 35.6 TFlops | 756.5 TFlops |
| BF16 | 35.6 TFlops | 756.5 TFlops |
| FP8 | — | 1513.0 TFlops |
| INT8 | 71.2 TFlops | 1513.0 TFlops |
FLOPS by Precision
Efficiency — TDP vs FP32 Performance
Analysis
Most Efficient
NVIDIA
H100
0.15 TFLOPS/W
Get Comparison Updates
New GPUs added weekly. Be the first to see how they compare.