Specifications Compared
| Spec | A100 SXM4 40GB | L40S |
|---|---|---|
| TDP | 400W | 350W |
| VRAM | 40 GB | 48 GB |
| CUDA Cores | 6,912 | 18,176 |
| FP8 (dense) | Not published | 733 TFLOPS |
| Memory Type | HBM2e | GDDR6 |
| Architecture | Ampere | Ada Lovelace |
| FP16 (dense) | 312 TFLOPS | 362.05 TFLOPS |
| Form Factors | SXM4 | PCIe |
| INT8 (dense) | 624 TOPS | 733 TOPS |
| Interconnect | NVLink, PCIe 4.0, InfiniBand | PCIe 4.0 |
| Tensor Cores | 432 | 568 |
| FP32 Performance | 19.5 TFLOPS | 91.6 TFLOPS |
| FP64 Performance | 9.7 TFLOPS | 1.4 TFLOPS |
| Memory Bandwidth | 2,039 GB/s | 864 GB/s |
| FP8 (with sparsity) | Not published | 1,466 TFLOPS |
| FP16 (with sparsity) | 624 TFLOPS | 733 TFLOPS |
| INT8 (with sparsity) | 1,248 TOPS | 1,466 TOPS |
Performance Analysis
The A100 SXM4 40GB delivers FP16 dense performance of 312 TFLOPS and FP32 performance of 19.5 TFLOPS while the L40S delivers FP16 dense performance of 362.05 TFLOPS and FP32 performance of 91.6 TFLOPS. This FP16 dense to FP16 dense comparison shows the L40S holds an advantage in throughput for mixed precision operations yet the A100 SXM4 40GB memory bandwidth of 2039 GB/s exceeds the L40S bandwidth of 864 GB/s which allows larger batch sizes during training sessions. When comparing FP16 with sparsity the A100 SXM4 40GB reaches 624 TFLOPS against the L40S figure of 733 TFLOPS so inference workloads that activate sparsity features see higher effective rates on the L40S.
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.
A100 SXM4 40GB
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| LeaderGPU | The Netherlands | 8 | $0.68 | $5.42 | Deploy |
| Vast.ai | Czechia, CZ | 1 | $0.73 | — | Deploy |
| ThunderCompute | USA | 2 | $1.09 | $2.18 | Deploy |
| Hyperstack | CANADA-1 | 1 | $1.35 | — | Deploy |
| Massed Compute | us-central-3 | 1 | $1.35 | — | Deploy |
11 providers in stock, 42 offers (cheapest per provider shown). All A100 SXM4 40GB offers, price history and alerts
L40S
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| Vast.ai | Slovenia, SI | 1 | $0.80 | — | Deploy |
| Massed Compute | us-central-2 | 8 | $0.97 | $7.76 | Deploy |
| QuantaCloud | us-midwest-2 | 1 | $1.09 | — | Deploy |
| RunPod | global | 1 | $1.09 | — | Deploy |
| Ori | lille-4 | 1 | $1.55 | — | Deploy |
6 providers in stock, 16 offers (cheapest per provider shown). All L40S offers, price history and alerts
Notify me when A100 SXM4 40GB drops below a price
One email when the cheapest in-stock on-demand price per GPU-hour falls below your threshold. Cheapest right now: $0.68/GPU-hr.
QuantaCloud
Comparing A100 providers? We broker across all of them.
Need 16+ A100s reserved for fine-tuning, simulation, or production inference? We quote volume pricing across multiple data center partners: one quote at partner rates, 24h turnaround.
When to Choose the A100 SXM4 40GB
The A100 SXM4 40GB suits deployments that require NVLink interconnect for multi GPU scaling and the higher memory bandwidth of 2039 GB/s supports scientific computing tasks that move large datasets. Its SXM4 form factor fits specific server designs that cannot accommodate standard PCIe cards and the 400W TDP aligns with power configurations already provisioned for Ampere based systems.
When to Choose the L40S
Its 350W TDP reduces power draw compared with the 400W figure of the A100 SXM4 40GB and the Ada Lovelace architecture provides higher dense FP16 throughput of 362.05 TFLOPS for inference pipelines.
Use Cases
The A100 SXM4 40GB provides 2039 GB/s memory bandwidth that sustains larger batch sizes during training runs.
The L40S achieves FP16 dense throughput of 362.05 TFLOPS which surpasses the 312 TFLOPS of the A100 SXM4 40GB.
Both GPUs deliver comparable INT8 dense performance of 624 TOPS on the A100 SXM4 40GB and 733 TOPS on the L40S for fine tuning workloads.
The L40S FP32 performance of 91.6 TFLOPS accelerates diffusion model operations that rely on single precision calculations.
The A100 SXM4 40GB NVLink interconnect enables direct GPU to GPU communication required in distributed scientific simulations.
Frequently Asked Questions
What memory types do the A100 SXM4 40GB and L40S use?▾
The A100 SXM4 40GB uses 40 GB HBM2e memory while the L40S uses 48 GB GDDR6 memory. These capacities represent the specific variant configurations compared here.
How do the form factors differ between these two GPUs?▾
The A100 SXM4 40GB uses an SXM4 form factor that supports NVLink while the L40S uses a PCIe form factor limited to PCIe 4.0 interconnect.
Which GPU has higher memory bandwidth?▾
The A100 SXM4 40GB provides 2039 GB/s memory bandwidth compared with 864 GB/s on the L40S. Both figures apply across all variants in their respective families.
What FP32 performance do these GPUs deliver?▾
The A100 SXM4 40GB delivers 19.5 TFLOPS in FP32 while the L40S delivers 91.6 TFLOPS in FP32. These values remain consistent for every variant within each family.
Do the GPUs support NVLink?▾
Only the A100 SXM4 40GB includes NVLink interconnect in addition to PCIe 4.0 and InfiniBand. The L40S supports PCIe 4.0 interconnect exclusively.
What are the TDP ratings for each GPU?▾
The A100 SXM4 40GB operates at a TDP of 400W while the L40S operates at a TDP of 350W. These power figures apply to all variants in each family.
Which is cheaper to rent, the A100 SXM4 40GB or the L40S?▾
Cloud rental prices for both the A100 SXM4 40GB 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 A100 SXM4 40GB have compared to the L40S?▾
The A100 SXM4 40GB has 40 GB of HBM2e memory. The L40S has 48 GB of GDDR6 memory.
Can I find A100 SXM4 40GB 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 A100 SXM4 40GB and the L40S?▾
The A100 SXM4 40GB uses the Ampere architecture (2020) while the L40S uses Ada Lovelace (2023). The L40S delivers 1.2x the dense FP16 throughput (362.05 vs 312 TFLOPS, both without sparsity) while the A100 SXM4 40GB has 2.4x 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 A100 SXM4 40GB 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 A100Every 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.