NVIDIA · Ada Lovelace Architecture
Rent NVIDIA L40S in the Cloud
Versatile data center GPU balancing AI compute with graphics capabilities. Strong for inference workloads.
VRAM
48 GB GDDR6
Bandwidth
864 GB/s
FP16
366.0 TFLOPS
FP32
91.6 TFLOPS
TDP
350W
Architecture
Ada Lovelace
Cheapest On-Demand
$0.55/hr
Average On-Demand
$0.95/hr
Providers
7
Compare NVIDIA L40S Cloud Pricing — 7 Providers
On-Demand
| Provider | Price / GPU / hr | Availability | Notes | Action |
|---|---|---|---|---|
Vast.ai
United States
|
$0.55/hr CHEAPEST | Available | Marketplace avg | Visit Provider |
Novita AI
United States
|
$0.55/hr CHEAPEST | Available | — | Visit Provider |
Latitude.sh
Brazil
|
$0.61/hr | Available | GPU VM | Visit Provider |
RunPod
United States
|
$0.86/hr | Available | Secure Cloud | Visit Provider |
Massed Compute
United States
|
$0.86/hr | Available | — | Visit Provider |
DigitalOcean
United States
|
$1.57/hr | Available | — | Visit Provider |
Vultr
United States
|
$1.67/hr | Available | — | Visit Provider |
Spot / Preemptible
| Provider | Price / GPU / hr | Availability | Notes | Action |
|---|---|---|---|---|
|
RunPod
United States
|
$0.79/hr CHEAPEST | Available | Community Cloud | Visit Provider |
Prices last verified: April 13, 2026
NVIDIA L40S Technical Specifications
Manufacturer
NVIDIA
Architecture
Ada Lovelace
VRAM
48 GB GDDR6
Memory Bandwidth
864 GB/s
FP16 (Tensor)
366.0 TFLOPS
FP32
91.6 TFLOPS
TDP
350W
Release Year
2023
Segment
Data center
Memory Type
GDDR6
Best For
Compare with Other GPUs
See how NVIDIA L40S stacks up against other popular cloud GPUs in specs, pricing, and availability.