Specifications Compared
| Spec | B200 NVL | L40S |
|---|---|---|
| TDP | 1000W | 350W |
| VRAM | 192 GB | 48 GB |
| CUDA Cores | 18,432 | 18,176 |
| FP4 (dense) | 9,000 TFLOPS | Not published |
| FP8 (dense) | 4,500 TFLOPS | 733 TFLOPS |
| Memory Type | HBM3e | GDDR6 |
| Architecture | Blackwell | Ada Lovelace |
| FP16 (dense) | 2,250 TFLOPS | 362.05 TFLOPS |
| Form Factors | NVL (PCIe card, NVLink bridged) | PCIe |
| INT8 (dense) | 4,500 TOPS | 733 TOPS |
| Interconnect | NVLink bridge, PCIe 6.0 | PCIe 4.0 |
| Tensor Cores | 576 | 568 |
| FP32 Performance | 75 TFLOPS | 91.6 TFLOPS |
| FP64 Performance | 37 TFLOPS | 1.4 TFLOPS |
| Memory Bandwidth | 8,000 GB/s | 864 GB/s |
| FP4 (with sparsity) | 18,000 TFLOPS | Not published |
| FP8 (with sparsity) | 9,000 TFLOPS | 1,466 TFLOPS |
| FP16 (with sparsity) | 4,500 TFLOPS | 733 TFLOPS |
| INT8 (with sparsity) | 9,000 TOPS | 1,466 TOPS |
Performance Analysis
The FP16 dense throughput of 2250 TFLOPS on the B200 NVL variant exceeds the 362.05 TFLOPS FP16 dense figure of the L40S variant by a factor derived directly from those two numbers. The same ratio holds when comparing FP16 with sparsity at 4500 TFLOPS versus 733 TFLOPS. FP32 throughput runs in the opposite direction with the L40S variant delivering 91.6 TFLOPS against the 75 TFLOPS of the B200 NVL variant. Memory bandwidth of 8000 GB/s shared across all B200 variants versus 864 GB/s shared across all L40S variants determines the largest sustainable batch sizes for models that fit inside 192 GB versus those constrained to 48 GB.
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.
B200 NVL
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| RunPod | global | 1 | $7.99 | — | Deploy |
| Vast.ai | , US | 8 | $9.00 | $72.00 | Deploy |
2 providers in stock, 3 offers (cheapest per provider shown). All B200 NVL 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-3 | 4 | $0.97 | $3.88 | Deploy |
| RunPod | global | 1 | $1.09 | — | Deploy |
| QuantaCloud | us-midwest-1 | 8 | $1.09 | $8.72 | Deploy |
| Lyceum | Europe | 2 | $1.99 | $3.98 | Deploy |
5 providers in stock, 18 offers (cheapest per provider shown). All L40S offers, price history and alerts
Notify me when B200 NVL drops below a price
One email when the cheapest in-stock on-demand price per GPU-hour falls below your threshold. Cheapest right now: $7.99/GPU-hr.
QuantaCloud
Comparing B-series options? Get one quote for all of them.
Skip the per-provider sales calls. Reserved and cluster B-series configurations from 16 to 1024+ GPUs with InfiniBand fabric, 3 to 12 month terms. One quote at partner rates, 24h turnaround.
When to Choose the B200 NVL
The B200 NVL variant suits deployments that require the full 192 GB HBM3e capacity and 8000 GB/s bandwidth to accommodate models whose working sets exceed 48 GB. Its NVLink bridge and PCIe 6.0 interconnect further support multi GPU scaling under the 1000 W TDP envelope of the Blackwell family.
When to Choose the L40S
The L40S variant fits installations where the 350 W TDP and PCIe 4.0 interface reduce power and cooling demands while the 48 GB GDDR6 memory remains sufficient. Its 91.6 TFLOPS FP32 rating provides an advantage for workloads that rely on that precision rather than the tensor figures shared by the Ada Lovelace family.
Use Cases
The 192 GB capacity and 2250 TFLOPS FP16 dense throughput of the B200 NVL variant support larger models than the 48 GB and 362.05 TFLOPS limits of the L40S variant.
The 8000 GB/s bandwidth shared by all B200 variants sustains higher token rates for models that fit inside 192 GB compared with the 864 GB/s bandwidth of the L40S variant.
The 4500 TFLOPS FP16 with sparsity figure of the B200 NVL variant accelerates gradient updates on models larger than those that fit in 48 GB on the L40S variant.
The 91.6 TFLOPS FP32 rating of the L40S variant exceeds the 75 TFLOPS of the B200 NVL variant for diffusion workloads that remain within the 48 GB memory limit.
Workloads that fit in 48 GB can use the lower 350 W TDP of the L40S variant while those needing 192 GB must select the B200 NVL variant.
Frequently Asked Questions
How do the FP16 dense figures compare between the two GPUs?▾
The B200 NVL variant delivers 2250 TFLOPS FP16 dense while the L40S variant delivers 362.05 TFLOPS FP16 dense. Both figures are family level values shared across all variants of each architecture.
What interconnects are used by each variant?▾
The B200 NVL variant uses an NVLink bridge together with PCIe 6.0 while the L40S variant uses PCIe 4.0. These interconnect details belong to the specific variants rather than the broader families.
What TDP values apply to the B200 NVL and L40S variants?▾
The B200 NVL variant operates at the 1000 W TDP shared by the Blackwell family while the L40S variant operates at the 350 W TDP shared by the Ada Lovelace family.
Which variant offers higher memory bandwidth?▾
All B200 variants share 8000 GB/s bandwidth while all L40S variants share 864 GB/s bandwidth. The difference therefore applies directly to the B200 NVL versus L40S comparison.
Which is cheaper to rent, the B200 NVL or the L40S?▾
Cloud rental prices for both the B200 NVL 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 B200 NVL have compared to the L40S?▾
The B200 NVL has 192 GB of HBM3e memory. The L40S has 48 GB of GDDR6 memory.
Can I find B200 NVL 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 B200 NVL and the L40S?▾
The B200 NVL uses the Blackwell architecture (2024) while the L40S uses Ada Lovelace (2023). The B200 NVL delivers 6.2x the dense FP16 throughput (2,250 vs 362.05 TFLOPS, both without sparsity) and 9.3x 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 B200 NVL 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 B200Every 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.