NVIDIA · Ampere · Data centre
NVIDIA A100
NVIDIA's Ampere data-centre GPU. Older than Hopper and without FP8, but still widely offered for training and inference.
01 / SPECIFICATIONS
What the hardware holds
- Memory
- 80 GB
- Memory bandwidth
- 2.04 TB/s
- FP8 dense
- Not published
- FP16 / BF16 dense
- 312 TFLOPS
- Variant
- A100 80GB SXM
- Memory type
- HBM2e
- Form factor
- SXM4
- Interconnect
- NVLink 600 GB/s
- Max power
- 400 W
Source notes for A100 80GB SXM
Datasheet lists FP16/BF16 Tensor '312 TFLOPS | 624 TFLOPS*' (* with sparsity); 312 is dense. No FP8 Tensor Core support (null). 400W standard configuration; HGX A100-80GB CTS SKU supports up to 500W. Bandwidth 2,039 GB/s. Also PCIe Gen4 64 GB/s.
Compute figures are dense, without sparsity. NVIDIA source for A100 80GB SXM Checked 8 Nov 2026.
02 / MEMORY FIT
How many do you need to hold a model?
| Model size | FP16 / BF16 | FP8 | INT4 |
|---|---|---|---|
| 8B parameters | 1 GPU16 GB of weights | 1 GPU8 GB of weights | 1 GPU4 GB of weights |
| 70B parameters | 3 GPUs140 GB of weights | 2 GPUs70 GB of weights | 1 GPU35 GB of weights |
| 405B parameters | 13 GPUs810 GB of weights | 7 GPUs405 GB of weights | 4 GPUs203 GB of weights |
GPUs = parameters × bytes per parameter × 1.2 headroom ÷ 80 GB, rounded up. This covers weights only. Long contexts, large batches, and training need much more memory.
03 / OFFERS
Reviewed on-demand rates
| Provider | Configuration | Status | Source | ||
|---|---|---|---|---|---|
 Runpod | A100 80GB PCIe1× A100 PCIe Pod · 80 GB | $1.79$1.79 ÷ 1 GPU | $1.791 GPU | Stock not verified | Rate card |
Per GPU-hour divides a machine's hourly price by its GPU count, so single GPUs and 8-GPU nodes compare fairly. The billed price is what one machine costs per hour. Storage, network transfer, and minimum terms can add to the bill.
04 / SELLERS
Who lists the A100
- RunpodGPU cloud · Self-serve · Reviewed rate
LambdaGPU cloud · Self-serve
CoreWeaveGPU cloud · Sales-led
CrusoeGPU cloud · Self-serve
Vast.aiMarketplace · Self-serve
Lambda, CoreWeave, Crusoe, Vast.ai list the A100 without a rate we have reviewed. Ask them for a quote, especially for clusters of 8 or more GPUs.
05 / COMPARE WITH