Specifications Compared
| Spec | H100 SXM5 | L40S |
|---|---|---|
| TDP | 700W | 350W |
| VRAM | 80 GB | 48 GB |
| CUDA Cores | 16,896 | 18,176 |
| FP8 (dense) | 1,979 TFLOPS | 733 TFLOPS |
| Memory Type | HBM3 | GDDR6 |
| Architecture | Hopper | Ada Lovelace |
| FP16 (dense) | 989 TFLOPS | 362.05 TFLOPS |
| Form Factors | SXM5 | PCIe |
| INT8 (dense) | 1,979 TOPS | 733 TOPS |
| Interconnect | NVLink, PCIe 5.0, InfiniBand | PCIe 4.0 |
| Tensor Cores | 528 | 568 |
| FP32 Performance | 67 TFLOPS | 91.6 TFLOPS |
| FP64 Performance | 34 TFLOPS | 1.4 TFLOPS |
| Memory Bandwidth | 3,350 GB/s | 864 GB/s |
| FP8 (with sparsity) | 3,958 TFLOPS | 1,466 TFLOPS |
| FP16 (with sparsity) | 1,979 TFLOPS | 733 TFLOPS |
| INT8 (with sparsity) | 3,958 TOPS | 1,466 TOPS |
Performance Analysis
The H100 SXM5 delivers 989 TFLOPS in dense FP16 compared with 362.05 TFLOPS in dense FP16 on the L40S. The H100 SXM5 also reaches 1979 TFLOPS in FP16 with sparsity against 733 TFLOPS in FP16 with sparsity on the L40S. FP32 performance stands at 67 TFLOPS for the H100 family and 91.6 TFLOPS for the L40S family. Higher memory bandwidth of 3350 GB/s on the H100 family supports larger batch sizes during training workloads than the 864 GB/s bandwidth on the L40S family. Dense FP8 figures show 1979 TFLOPS on the H100 family versus 733 TFLOPS on the L40S family.
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.
H100 SXM5
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| QuantaCloud | us-midwest-2 | 1 | $2.59 | — | Deploy |
| Massed Compute | us-central-3 | 1 | $2.73 | — | Deploy |
| Ori | dallas-2 | 1 | $2.90 | — | Deploy |
| LeaderGPU | The Netherlands | 8 | $3.59 | $28.75 | Deploy |
| Vast.ai | Czechia, CZ | 2 | $4.23 | $8.47 | Deploy |
9 providers in stock, 20 offers (cheapest per provider shown). All H100 SXM5 offers, price history and alerts
L40S
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| Massed Compute | us-central-3 | 1 | $0.97 | — | Deploy |
| RunPod | global | 1 | $1.09 | — | Deploy |
| QuantaCloud | us-midwest-2 | 2 | $1.09 | $2.18 | Deploy |
| Lyceum | Europe | 1 | $2.19 | — | Deploy |
4 providers in stock, 5 offers (cheapest per provider shown). All L40S offers, price history and alerts
Notify me when H100 SXM5 drops below a price
One email when the cheapest in-stock on-demand price per GPU-hour falls below your threshold. Cheapest right now: $2.59/GPU-hr.
QuantaCloud
Comparing H-series providers? We broker across all of them.
Hopper stock changes by the hour and prices differ by provider. If you need 16+ GPUs reserved or a cluster in the next 90 days, we quote H-series or B300 inventory at partner rates: one quote, 24h turnaround.
When to Choose the H100 SXM5
The H100 SXM5 suits workloads that require the 80 GB HBM3 capacity and 3350 GB/s bandwidth for large model training. Its NVLink interconnect enables scaling across multiple units in dense configurations where the L40S PCIe 4.0 limit does not apply.
When to Choose the L40S
The L40S suits workloads that benefit from its 91.6 TFLOPS FP32 rating and 350W TDP in power constrained environments. Its 48 GB GDDR6 configuration fits inference tasks that do not require the 80 GB capacity of the H100 SXM5.
Use Cases
The H100 SXM5 supplies 80 GB HBM3 and 989 TFLOPS dense FP16 that exceed the 48 GB and 362.05 TFLOPS on the L40S.
The H100 SXM5 provides 1979 TFLOPS FP16 with sparsity and NVLink scaling that exceed the 733 TFLOPS and PCIe 4.0 limit on the L40S.
The H100 SXM5 offers 80 GB HBM3 capacity and 3350 GB/s bandwidth that support larger fine tuning batches than the L40S.
The L40S delivers 91.6 TFLOPS FP32 and 350W TDP that fit diffusion workloads without needing the higher TDP of the H100 SXM5.
The L40S provides 91.6 TFLOPS FP32 that exceeds the 67 TFLOPS on the H100 family for precision sensitive simulations.
Frequently Asked Questions
How does FP16 performance compare between these GPUs?▾
The H100 family delivers 989 TFLOPS dense FP16 and 1979 TFLOPS FP16 with sparsity. The L40S family delivers 362.05 TFLOPS dense FP16 and 733 TFLOPS FP16 with sparsity.
What form factor and interconnect does the L40S use?▾
The L40S variant uses a PCIe form factor with PCIe 4.0 interconnect. The H100 SXM5 variant uses an SXM5 form factor with NVLink support.
What is the TDP difference between the H100 and L40S families?▾
Every H100 family member has a TDP of 700W. Every L40S family member has a TDP of 350W.
Which GPU offers higher FP32 throughput?▾
The L40S family offers 91.6 TFLOPS FP32. The H100 family offers 67 TFLOPS FP32.
Which is cheaper to rent, the H100 SXM5 or the L40S?▾
Cloud rental prices for both the H100 SXM5 and L40S 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 H100 SXM5 have compared to the L40S?▾
The H100 SXM5 has 80 GB of HBM3 memory. The L40S has 48 GB of GDDR6 memory.
Can I find H100 SXM5 and L40S 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 H100 SXM5 and the L40S?▾
The H100 SXM5 uses the Hopper architecture (2022) while the L40S uses Ada Lovelace (2023). The H100 SXM5 delivers 2.7x the dense FP16 throughput (989 vs 362.05 TFLOPS, both without sparsity) and 3.9x 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 H100 SXM5 and the L40S. 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 H100Every current offer by provider, daily price history and a price alert.
- Rent L40SEvery 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.