Flopper.io GPU Comparison Sheet
NVIDIA A100 PCIe 80GB vs NVIDIA GeForce RTX 3090 Ti 24GB
Side by side datacenter GPU specifications from the Flopper.io database.
Overview
| NVIDIA A100 PCIe 80GB | NVIDIA GeForce RTX 3090 Ti 24GB | |
|---|---|---|
| FP32 TFLOPS | 19.5 | 40.0 |
| VRAM | 80 GB | 24 GB |
| Memory Bandwidth | 1.9 TB/s | 1.0 TB/s |
| TDP | 300 W | 450 W |
Performance by Precision
| Precision | NVIDIA A100 PCIe 80GB peak TFLOPS (dense) | NVIDIA GeForce RTX 3090 Ti 24GB peak TFLOPS (dense) |
|---|---|---|
| INT8 | 624.0 | 80.0 |
| INT4 | 1248.0 | — |
| FP16 | 312.0 | 40.0 |
| BF16 | 312.0 | 40.0 |
| TF32 | 156.0 | 40.0 |
| FP32 | 19.5 | 40.0 |
| FP64 | 9.7 | 0.6 |
Specifications
| NVIDIA A100 PCIe 80GB | NVIDIA GeForce RTX 3090 Ti 24GB | |
|---|---|---|
| Architecture | Ampere | Ampere |
| Process Node | — | 8nm |
| Launch Year | 2021 | 2022 |
| Form Factor | PCIe | PCIe |
| Memory Type | HBM2e | — |
| Interface Width | 5120-bit | — |
| Interconnect | — | — |
| TDP | 300 W | 450 W |
| FP32 Efficiency | 0.07 TFLOPS/W | 0.09 TFLOPS/W |