NVIDIA · Ada Lovelace
L40S 48 GB rental prices and specifications
A 48 GB PCIe datacenter GPU aimed at generative-AI inference, fine-tuning, rendering, and media workloads without NVLink or MIG support.
Profile checked · Primary source ↗
- Memory
- 48 GB
- Bandwidth
- 864 GB/s
- Current offers
- 0
- FP16 dense
- 362.05 TFLOPS
- BF16 dense
- 362.05 TFLOPS
- FP8 dense
- 733 TFLOPS
- Interconnect
- PCIe Gen4 x16: 64 GB/s bidirectional; NVLink is not supported.
Compare L40S 48 GB cloud rentals
Use the exact 48 GB GPU profile when comparing rental quotes. Tables separate currencies, billing periods and rental terms; the required GPU count determines the whole-instance compute cost. Storage, transfer, taxes and other charges can be additional.
Source checks show when each price was actually observed. Historical quotes remain available separately and do not enter current price comparisons. Explore the sourced GPU rental price index or check a rental against your model requirements.
Uses the explicit dense column of the current NVIDIA specification table (including its reported FP16/BF16 362.05 TFLOPS); sparse numbers are not substituted or recomputed. · Specification source 1 ↗
Recently checked listings
Current rental offers
| GPU / provider | Instance price / offer | Terms / location |
|---|---|---|
| No offers match these filters. Try another billing period or currency. | ||
Previously observed offers
69 retained quotes with an old check, a different allocation, or unavailable status. These are excluded from current prices and capacity counts.
Show previously observed offers →Measured performance
Inference benchmarks
| Model | GPU setup | Engine | Quant. | Workload | Reported tok/s | Evidence | Max-load capacity scenarioUSD / million output tokens |
|---|---|---|---|---|---|---|---|
| Qwen3-30B-A3B | L40S 48 GB1 GPU | SGLangMeasured SGLang build undisclosed; AIPerf v0.11.0 | FP8Qwen/Qwen3-30B-A3B-FP8; measured revision undisclosed | 1024 → 256input → output | 897.4 | Source ↗Published Aug 29, 2026 | —No recent non-reference on-demand USD offer matches this GPU variant and count.One NVIDIA L40S 48 GB GPU; author-run SGLang service |
Output throughput is aggregate maximum-load performance, not single-user generation speed. Workloads show input → output tokens. Compare rows only when the model, GPU count, quantization, engine, and workload match.
Capacity scenarios combine an exact GPU variant/count rental with published output throughput from a different system. Throughput was not measured on the selected rental. Taxes, extras, idle time and minimum billing are excluded; these are not measured operating costs. Calculation guide.