Specifications Compared
| Spec | L40S | RTX 4090 |
|---|---|---|
| TDP | 350W | 450W |
| VRAM | 48 GB | 24 GB |
| CUDA Cores | 18,176 | 16,384 |
| FP8 (dense) | 733 TFLOPS | 330.3 TFLOPS |
| Memory Type | GDDR6 | GDDR6X |
| Architecture | Ada Lovelace | Ada Lovelace |
| FP16 (dense) | 362.05 TFLOPS | 165.2 TFLOPS |
| Form Factors | PCIe | PCIe |
| INT8 (dense) | 733 TOPS | 660.6 TOPS |
| Interconnect | PCIe 4.0 | PCIe 4.0 |
| Tensor Cores | 568 | 512 |
| FP32 Performance | 91.6 TFLOPS | 82.6 TFLOPS |
| FP64 Performance | 1.4 TFLOPS | 1.3 TFLOPS |
| Memory Bandwidth | 864 GB/s | 1,008 GB/s |
| FP8 (with sparsity) | 1,466 TFLOPS | 660.6 TFLOPS |
| FP16 (with sparsity) | 733 TFLOPS | 330.4 TFLOPS |
| INT8 (with sparsity) | 1,466 TOPS | 1,321.2 TOPS |
Performance Analysis
Dense FP16 throughput on the L40S reaches 362.05 TFLOPS while the RTX 4090 reaches 165.2 TFLOPS so training runs that rely on dense matrix operations complete more iterations per hour on the L40S. The same pattern holds for dense FP8 where the L40S lists 733 TFLOPS against 330.3 TFLOPS on the RTX 4090. Memory bandwidth of 864 GB/s on the L40S versus 1008 GB/s on the RTX 4090 limits maximum batch size on the L40S during inference yet the doubled VRAM capacity offsets this constraint for models exceeding 24 GB. Sparse FP16 figures of 733 TFLOPS on the L40S and 330.4 TFLOPS on the RTX 4090 maintain the same ratio when sparsity is enabled.
Current On-Demand Offers
Cheapest secure on-demand offer per provider that is reported in stock right now, per GPU per hour; multi-GPU instances show the whole-instance price alongside. Deploy opens the DeployGPU console for providers on the platform, otherwise the provider's own site. Stock is rechecked every minute.
L40S
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| Massed Compute | us-central-3 | 4 | $0.97 | $3.88 | Deploy |
| QuantaCloud | us-midwest-1 | 4 | $1.09 | $4.36 | Deploy |
| RunPod | global | 1 | $1.09 | — | Deploy |
| Lyceum | Europe | 1 | $1.99 | — | Deploy |
4 providers in stock, 17 offers (cheapest per provider shown). All L40S offers, price history and alerts
Notify me when L40S drops below a price
One email when the cheapest in-stock on-demand price per GPU-hour falls below your threshold. Cheapest right now: $0.97/GPU-hr.
QuantaCloud
Comparing providers? We broker across all of them.
Stop tab-switching between pricing pages. Tell us what you need, 16+ GPUs reserved or cluster capacity, and we return one quote at partner rates within 24 hours.
When to Choose the L40S
The L40S suits LLM training workloads that require 48 GB VRAM to hold full precision weights and optimizer states. Its 91.6 TFLOPS FP32 rating and 350W TDP further support continuous scientific computing jobs where power draw and memory headroom determine job completion rates.
When to Choose the RTX 4090
Its 660.6 TOPS dense INT8 rating also supports latency sensitive Stable Diffusion batches where memory transfer speed outweighs total capacity.
Use Cases
The L40S supplies 48 GB VRAM and 362.05 TFLOPS dense FP16 which accommodate larger batch sizes than the 24 GB and 165.2 TFLOPS dense FP16 on the RTX 4090.
The RTX 4090 provides 1008 GB/s bandwidth which reduces token latency for models that remain under the 24 GB limit compared with 864 GB/s on the L40S.
The L40S 48 GB capacity and 733 TFLOPS dense FP8 rating support full parameter updates without offloading while the RTX 4090 is restricted to 24 GB.
The RTX 4090 1008 GB/s bandwidth accelerates image generation steps for diffusion models that fit in 24 GB compared with the L40S bandwidth.
The L40S 91.6 TFLOPS FP32 and 350W TDP enable sustained double precision workloads with greater memory capacity than the RTX 4090 82.6 TFLOPS FP32.
Frequently Asked Questions
What is the FP32 performance of the L40S versus the RTX 4090?▾
The L40S reaches 91.6 TFLOPS FP32 and the RTX 4090 reaches 82.6 TFLOPS FP32. The higher rating on the L40S benefits scientific computing kernels that rely on single precision.
Which GPU has higher memory bandwidth?▾
The RTX 4090 lists 1008 GB/s while the L40S lists 864 GB/s. Higher bandwidth on the RTX 4090 improves data movement for bandwidth bound inference workloads.
How do the dense FP16 figures compare?▾
The L40S delivers 362.05 TFLOPS dense FP16 and the RTX 4090 delivers 165.2 TFLOPS dense FP16. The L40S therefore processes twice the dense half precision operations per second.
What TDP values are listed for each card?▾
The L40S TDP is 350W and the RTX 4090 TDP is 450W. Lower power on the L40S reduces cooling and electricity requirements in multi GPU servers.
Do both GPUs use the same interconnect?▾
Both the L40S and RTX 4090 use PCIe 4.0. This shared interface means data transfer rates between host and device remain equivalent.
Which is cheaper to rent, the L40S or the RTX 4090?▾
Cloud rental prices for both the L40S and RTX 4090 vary by provider, configuration, and availability. This page shows live pricing from 25+ providers updated every 60 seconds. Scroll to the Live Cloud Pricing section to compare current rates.
How much VRAM does the L40S have compared to the RTX 4090?▾
The L40S has 48 GB of GDDR6 memory. The RTX 4090 has 24 GB of GDDR6X memory.
Can I find L40S and RTX 4090 GPUs available to rent right now?▾
Yes. This page shows real-time availability across 25+ cloud GPU providers. The Live Cloud Pricing section displays only in-stock offers with current pricing.
What is the main difference between the L40S and the RTX 4090?▾
The L40S uses the Ada Lovelace architecture (2023) while the RTX 4090 uses Ada Lovelace (2022). The L40S delivers 2.2x the dense FP16 throughput (362.05 vs 165.2 TFLOPS, both without sparsity) while the RTX 4090 has 1.2x the memory bandwidth of the L40S.
Rent these GPUs
Each GPU page lists every current on-demand offer by provider, per GPU-hour, updated every minute.
Related comparisons
How this page is made
- Specifications come from the NVIDIA, AMD and Intel datasheets for the L40S and the RTX 4090. Dense and sparse throughput are listed separately.
- Prices and availability are live: every offer shown is an in-stock, on-demand listing from a provider we track, rechecked every minute.
- The written comparison (overview, performance notes, when to choose each card, verdict, use cases and FAQ) was drafted with an AI model from the specification table above and passed an automated check that rejects any figure not in that table. It was last generated on . It contains no prices; those are always read live.
- Read how we collect the data, or report an error on this page.
Next steps
- Rent L40SEvery current offer by provider, daily price history and a price alert.
- Rent RTX 4090Every current offer by provider, daily price history and a price alert.
- GPU price indexHow on-demand prices for the major GPUs have moved, updated daily.
- All GPUsEvery GPU we track, with current lows and provider counts.