NVIDIA · Data centre
L40S
Rented by 3 providers. 2 exact variants are priced separately, because they are different silicon.
Best published rate
$0.790
per GPU / hour
RunPod · 48 GB
RunPod · 48 GB
Every published rate
cheapest first · USD per GPU per hour| Provider | What they call it | Their own price, and the arithmetic | Per GPU-hour |
|---|---|---|---|
| RunPod
48 GB |
L40S
NVIDIA L40S |
$0.7900 per hour = $0.7900 per GPU-hour
region Secure Cloud · 1 offer Medium stock |
$0.7900
$2.18 per PFLOP-hr |
| Vultr
48 GB |
L40S
vbm-64c-2048gb-8-l40-gpu |
$13.37 per hour ÷ 8 GPUs = $1.6710 per GPU-hour
regions not published in this feed · 4 offers |
$1.6710
$2.18 per PFLOP-hr |
| Scaleway
48 GB |
L40S
L40S-8-48G |
0.1959888 EUR per 1 minute → 11.7593 EUR/hr × 1.1699 EUR/USD ÷ 8 GPUs = $1.7197 per GPU-hour
region fr-par-2 · 4 offers |
$1.7197
$2.18 per PFLOP-hr |
| Scaleway
form factor and memory not stated |
L40S
EM-T220E-L40S |
4.11 EUR per 1 hour → 4.1100 EUR/hr × 1.1699 EUR/USD ÷ 2 GPUs = $2.4041 per GPU-hour
region fr-par-2 · 1 offer |
$2.4041 |
The silicon
vendor-published| Variant | Architecture | Memory | FP16 dense | Board power |
|---|---|---|---|---|
| L40S | Ada Lovelace 2023 | 48 GB GDDR6 | 362.05 TFLOPS | 350 W |
Vendor-published figures. FP16 throughput is dense tensor throughput with no sparsity assumed; vendors often headline the sparse figure, which is twice as large.
Vendor page ↗
1 of this chip's 2 variants carry no vendor specification here — they are priced all the same, and simply say nothing where a figure would go.
What they call it
Providers write the same hardware down differently. Every name below is one this index has seen for this chip; nothing was renamed to make the table tidier.
L40S