Flopper.io System Specification Sheet
August 10, 2026
flopper.io/system/nvidia-hgx-a100-8-gpuNVIDIA HGX A100
8x A100 Ampere baseboard 2020
Overview
2.50
BF16 PFLOPS Dense
0.6 TB
Aggregate Memory
16.3 TB/s
Aggregate Bandwidth
6.0 kW
Aggregate Power
Performance Metrics
| Precision | Peak (PFLOPS) | Per GPU (TFLOPS) | TFLOPS/W |
|---|---|---|---|
| FP64 | 0.1 | 9.7 | 12.93 |
| FP32 | 0.2 | 19.5 | 26.00 |
| TF32 | 1.2 | 156.0 | 208.00 |
| FP16 | 2.5 | 312.0 | 416.00 |
| BF16 | 2.5 | 312.0 | 416.00 |
| INT8 | 5.0 | 624.0 | 832.00 |
All figures are dense. Vendors commonly headline the sparse number, which is twice the dense one.
Configuration
- Accelerator
- A100 SXM4 80GB
- Count
- 8
- Memory per GPU
- 80 GB
- Bandwidth per GPU
- 2.0 TB/s
Power & Efficiency
- Aggregate Power
- 6.0 kW
- TDP per GPU
- 400 W
- BF16 Efficiency
- 416.00 TFLOPS/W
- Aggregate Bandwidth
- 16.3 TB/s
System Details
Vendor: NVIDIA
Family: HGX
Architecture: Ampere
Form Factor: baseboard
Release Year: 2020
Process Node: 7nm
Interconnect
NVSwitch + NVLink
Notes
HGX A100 platform for cloud providers