NVIDIA A100 80GB SXM
Data center GPU · Ampere
Summary
The most powerful of four A100 configurations, this GPU is well-suited to multi-GPU training or other workloads where GPU-to-GPU communication is important.
Launched
Q3 2021
VRAM
80 GB
Mem. bandwidth
2,039 GB/s
On-demand from
Tech specs
VRAM 80 GB HBM2e
Memory bandwidth 2,039 GB/s
Interface SXM
CUDA cores 6,912
Tensor cores 432 (Gen 3)
TDP 400 W
Supported data types
FP64FP32FP16BF16INT8INT4
Cloud rental prices
16 providers · available in 29 countries · lowest price per GPU, per hour
On-demand
$0.58 – $5.07 /GPU/h
Reserved
$1.36 – $2.35 /GPU/h
Spot
$0.63 – $1.21 /GPU/h
| Provider | On-demand | Reserved | Spot |
|---|---|---|---|
 AceCloud | — | $1.63 | — |
 AWS | $3.43 | $1.46 | — |
| | $3.40 | $1.36 | $0.75 |
 CoreWeave | $2.70 | — | $1.21 |
 Crusoe | $2.30 | — | — |
 Denvr | $0.58 | — | — |
| | $5.07 | — | $0.65 |
 Hyperstack | $1.60 | $1.36 | — |
 Lambda | $2.79 | — | — |
 Massed Compute | $1.38 | — | — |
 Oracle | $4.00 | — | — |
 OVHcloud | $3.13 | — | — |
 Runpod | $1.49 | — | — |
| | $1.82 | — | — |
 Verda | $1.79 | $1.65 | $0.63 |
 Vultr | $2.80 | — | — |
Models that fit in VRAM
Open-weights models that fit in VRAM. Estimated using 🤗 accelerate, plus approximation for up to 8K context.
Media
Peak theoretical performance
INT8 Tensor Core 624 TOPS
BF16 Tensor Core 312 TFLOPS
FP16 Tensor Core 312 TFLOPS
TF32 Tensor Core 156 TFLOPS
FP32 19.5 TFLOPS
FP64 9.7 TFLOPS
FP64 Tensor Core 19.5 TFLOPS
Performance figures assume no sparsity; in cases where only sparse performance figures are published by the manufacturer, these are halved to give approximate dense performance.
Resources
Detailed documentation from the manufacturer.
Similar GPUs
Other accelerators you might compare.