Specifications Compared
| Spec | L4 | L40 |
|---|---|---|
| TDP | 72W | 300W |
| VRAM | 24 GB | 48 GB |
| CUDA Cores | 7,424 | 18,176 |
| FP8 (dense) | 242 TFLOPS | 362 TFLOPS |
| Memory Type | GDDR6 | GDDR6 |
| Architecture | Ada Lovelace | Ada Lovelace |
| FP16 (dense) | 121 TFLOPS | 181.05 TFLOPS |
| Form Factors | PCIe | PCIe |
| INT8 (dense) | 242 TOPS | 362 TOPS |
| Interconnect | PCIe 4.0 | PCIe 4.0 |
| Tensor Cores | 232 | 568 |
| FP32 Performance | 30.3 TFLOPS | 90.5 TFLOPS |
| FP64 Performance | 0.5 TFLOPS | Not published |
| Memory Bandwidth | 300 GB/s | 864 GB/s |
| FP8 (with sparsity) | 485 TFLOPS | 724 TFLOPS |
| FP16 (with sparsity) | 242 TFLOPS | 362.1 TFLOPS |
| INT8 (with sparsity) | 485 TOPS | 724 TOPS |
Performance Analysis
The FP16 dense figure of 121 TFLOPS on the L4 exceeds its FP32 value of 30.3 TFLOPS by a factor of four while the L40 shows an FP16 dense value of 181.05 TFLOPS against its FP32 value of 90.5 TFLOPS by a factor of two. This ratio indicates the L4 favors mixed precision operations more than the L40 in training scenarios that alternate between FP16 dense and FP32. Memory bandwidth of 300 GB/s on the L4 versus 864 GB/s on the L40 limits maximum batch sizes on the L4 to roughly one third those possible on the L40 during large model inference. FP16 with sparsity comparisons show 242 TFLOPS on the L4 against 362.1 TFLOPS on the L40 while FP8 with sparsity reaches 485 TFLOPS versus 724 TFLOPS.
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.
L4
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| RunPod | global | 1 | $0.59 | — | Deploy |
| Scaleway | Warsaw, Poland (WAW2) | 2 | $0.89 | $1.77 | Deploy |
2 providers in stock, 4 offers (cheapest per provider shown). All L4 offers, price history and alerts
L40
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| ThunderCompute | USA | 2 | $0.79 | $1.58 | Deploy |
| Massed Compute | us-central-3 | 1 | $0.86 | — | Deploy |
| QuantaCloud | us-midwest-2 | 4 | $0.94 | $3.75 | Deploy |
| Hyperstack | CANADA-1 | 1 | $1.00 | — | Deploy |
4 providers in stock, 15 offers (cheapest per provider shown). All L40 offers, price history and alerts
Notify me when L4 drops below a price
One email when the cheapest in-stock on-demand price per GPU-hour falls below your threshold. Cheapest right now: $0.59/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 L4
The L4 suits deployments where TDP must remain at 72W and memory needs stay within 24 GB. PCIe interconnect at 4.0 supports these compact installations without exceeding power limits of edge servers. FP32 performance of 30.3 TFLOPS meets requirements for lighter scientific computing tasks that do not demand the 90.5 TFLOPS of the L40.
When to Choose the L40
The L40 fits scenarios that require 48 GB of memory and 864 GB/s bandwidth to handle larger batches. FP16 dense performance of 181.05 TFLOPS accelerates training and inference compared with the 121 TFLOPS of the L4. Higher TDP of 300W remains acceptable in data center racks that prioritize throughput over efficiency.
Use Cases
The L40 supplies 181.05 TFLOPS FP16 dense and 48 GB memory that exceed the L4 values for sustained training runs.
The L40 provides 362 TFLOPS FP8 dense and 864 GB/s bandwidth that allow larger batch sizes than the L4.
The L40 FP16 with sparsity at 362.1 TFLOPS supports fine-tuning workloads more efficiently than the L4 at 242 TFLOPS.
Both GPUs deliver adequate FP16 dense performance though the L40 handles higher resolution batches with 48 GB memory.
The L4 at 72W TDP and 30.3 TFLOPS FP32 meets requirements where power limits preclude the 300W L40.
Frequently Asked Questions
What is the memory difference between L4 and L40?▾
The L4 contains 24 GB GDDR6 while the L40 contains 48 GB GDDR6. Bandwidth measures 300 GB/s on the L4 and 864 GB/s on the L40.
How does FP16 performance compare on L4 versus L40?▾
FP16 dense reaches 121 TFLOPS on the L4 and 181.05 TFLOPS on the L40. FP16 with sparsity measures 242 TFLOPS on the L4 and 362.1 TFLOPS on the L40.
What TDP values apply to each GPU?▾
The L4 lists a TDP of 72W. The L40 lists a TDP of 300W. Both use PCIe form factors.
Which GPU offers higher FP32 performance?▾
The L40 reaches 90.5 TFLOPS FP32. The L4 reaches 30.3 TFLOPS FP32.
Do both GPUs support the same interconnect?▾
Both the L4 and L40 use PCIe 4.0 interconnect. Form factors remain PCIe for each model.
How do INT8 figures differ?▾
INT8 dense measures 242 TOPS on the L4 and 362 TOPS on the L40. INT8 with sparsity measures 485 TOPS on the L4 and 724 TOPS on the L40.
Which is cheaper to rent, the L4 or the L40?▾
Cloud rental prices for both the L4 and L40 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 L4 have compared to the L40?▾
The L4 has 24 GB of GDDR6 memory. The L40 has 48 GB of GDDR6 memory.
Can I find L4 and L40 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 L4 and the L40?▾
The L4 uses the Ada Lovelace architecture (2023) while the L40 uses Ada Lovelace (2023). The L40 delivers 1.5x the dense FP16 throughput (181.05 vs 121 TFLOPS, both without sparsity) and 2.9x the memory bandwidth of the L4.
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 L4 and the L40. 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 L4Every current offer by provider, daily price history and a price alert.
- Rent L40Every 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.