A100 vs L4
The A100 delivers 2.6× the FP16 compute of the L4 while costing 6.6× more on average, making the L4 157% more cost-efficient per TFLOP.
Spec Comparison
Spec
A100
L4
Architecture
Ampere
Ada Lovelace
VRAM
80 GB
24 GB
FP16 TFLOPS
312
122
Memory Bandwidth
2,000 GB/s
300 GB/s
MSRP
$10,000
$2,500
Used/Market Price
$7,000
$1,800
Live Pricing
Cost Efficiency
Metric
A100
L4
$/TFLOP/hr
$0.0266
$0.0104
$/GB-VRAM/hr
$0.1038
$0.0526
Lower $/TFLOP and $/GB-VRAM is better. Based on median hourly pricing.
Provider Availability
Only on L4 (5)
A100 Cloud Pricing
363 instances| Provider | GPUs | VRAM | Region | Type | Price/hr | |
|---|---|---|---|---|---|---|
| CHEAPESTVast.ai | 1x | 40 GB | EU-Central | Spot | $0.0800 | Rent |
| Vast.ai | 1x | 40 GB | EU-Central | On-Demand | $0.0934 | Rent |
| Vultr | 1x | 4 GB | ewr | On-Demand | $0.123 | Rent |
| Vultr | 1x | 4 GB | fra | On-Demand | $0.123 | Rent |
| Vultr | 1x | 4 GB | sjc | On-Demand | $0.123 | Rent |
| Vultr | 1x | 4 GB | nrt | On-Demand | $0.123 | Rent |
| Vultr | 1x | 8 GB | ewr | On-Demand | $0.247 | Rent |
| Vultr | 1x | 8 GB | nrt | On-Demand | $0.247 | Rent |
| Vultr | 1x | 8 GB | blr | On-Demand | $0.247 | Rent |
| Verda | 1x | 40 GB | EU-Central | Spot | $0.252 | Rent |
L4 Cloud Pricing
374 instances| Provider | GPUs | VRAM | Region | Type | Price/hr | |
|---|---|---|---|---|---|---|
| CHEAPESTAWS | 1x | 24 GB | US-East | Spot | $0.188 | Rent |
| GCP | 1x | 24 GB | Asia-Pacific | Spot | $0.192 | Rent |
| RunPod | 1x | 24 GB | EUR-IS-1 | Spot | $0.220 | Rent |
| RunPod | 1x | 24 GB | US-MO-2 | Spot | $0.220 | Rent |
| GCP | 1x | 24 GB | Asia-Pacific | Spot | $0.229 | Rent |
| AWS | 1x | 24 GB | US-East | Spot | $0.250 | Rent |
| AWS | 1x | 24 GB | EU-West | Spot | $0.261 | Rent |
| GCP | 1x | 24 GB | Asia-Pacific | Spot | $0.265 | Rent |
| AWS | 1x | 24 GB | Asia-Pacific | Spot | $0.272 | Rent |
| AWS | 1x | 24 GB | Asia-Pacific | Spot | $0.274 | Rent |