Specifications Compared
| Spec | H200 NVL | RTX 3090 |
|---|---|---|
| TDP | 700W | 350W |
| VRAM | 141 GB | 24 GB |
| CUDA Cores | 16,896 | 10,496 |
| FP8 (dense) | 1,979 TFLOPS | Not published |
| Memory Type | HBM3e | GDDR6X |
| Architecture | Hopper | Ampere |
| FP16 (dense) | 989 TFLOPS | 71 TFLOPS |
| Form Factors | NVL (PCIe card, NVLink bridged) | PCIe |
| INT8 (dense) | 1,979 TOPS | 284 TOPS |
| Interconnect | NVLink bridge, PCIe 5.0 | NVLink, PCIe 4.0 |
| Tensor Cores | 528 | 328 |
| FP32 Performance | 67 TFLOPS | 35.6 TFLOPS |
| FP64 Performance | 34 TFLOPS | Not published |
| Memory Bandwidth | 4,800 GB/s | 936 GB/s |
| FP8 (with sparsity) | 3,958 TFLOPS | Not published |
| FP16 (with sparsity) | 1,979 TFLOPS | 142 TFLOPS |
| INT8 (with sparsity) | 3,958 TOPS | 568 TOPS |
Performance Analysis
The H200 NVL family delivers 989 TFLOPS FP16 dense and 67 TFLOPS FP32 while the RTX 3090 family delivers 71 TFLOPS FP16 dense and 35.6 TFLOPS FP32. Memory bandwidth stands at 4800 GB/s for the H200 NVL family and 936 GB/s for the RTX 3090 family allowing larger batch sizes before memory becomes the limit. Sparse FP16 figures of 1979 TFLOPS on the H200 NVL family and 142 TFLOPS on the RTX 3090 family follow the same pattern 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.
H200 NVL
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| Vast.ai | Czechia, CZ | 2 | $4.92 | $9.84 | Deploy |
| RunPod | global | 1 | $5.29 | — | Deploy |
| VERDA | FIN-HEL | 2 | $5.60 | $11.20 | Deploy |
| Lyceum | Europe | 2 | $6.19 | $12.38 | Deploy |
| LeaderGPU | The Netherlands | 1 | $6.22 | — | Deploy |
5 providers in stock, 9 offers (cheapest per provider shown). All H200 NVL offers, price history and alerts
RTX 3090
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| Vast.ai | Finland, FI | 4 | $0.27 | $1.07 | Deploy |
| LeaderGPU | The Netherlands | 8 | $0.29 | $2.29 | Deploy |
2 providers in stock, 11 offers (cheapest per provider shown). All RTX 3090 offers, price history and alerts
Notify me when H200 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: $4.92/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 H200 NVL
The H200 NVL variant suits workloads that require the 141 GB memory capacity together with 4800 GB/s bandwidth and 989 TFLOPS FP16 dense throughput. Large model training and inference jobs that exceed 24 GB per card benefit directly from these shared family figures and the NVL form factor interconnect.
When to Choose the RTX 3090
The RTX 3090 variant fits smaller scale inference or fine tuning jobs where 24 GB GDDR6X memory and 350 W TDP suffice. Projects that operate within the 71 TFLOPS FP16 dense limit and do not need the higher 700 W TDP of the H200 NVL family gain from the lower power draw and PCIe form factor.
Use Cases
The 141 GB memory capacity supports models that exceed the 24 GB limit of the RTX 3090 variant while 989 TFLOPS FP16 dense provides higher throughput.
Batch sizes scale with 4800 GB/s bandwidth and 1979 TFLOPS FP16 with sparsity on the H200 NVL family.
The 141 GB capacity and 67 TFLOPS FP32 allow larger context windows than the 24 GB and 35.6 TFLOPS FP32 of the RTX 3090 variant.
The 24 GB GDDR6X memory and 350 W TDP meet typical diffusion workloads without requiring the higher power or capacity of the H200 NVL variant.
FP32 at 67 TFLOPS combined with 4800 GB/s bandwidth accelerates simulations beyond the 35.6 TFLOPS and 936 GB/s of the RTX 3090 variant.
Frequently Asked Questions
What is the FP16 dense performance of each GPU family?▾
The H200 NVL family lists 989 TFLOPS FP16 dense. The RTX 3090 family lists 71 TFLOPS FP16 dense. Both figures apply to every variant within each family.
Which variant uses more power?▾
The H200 NVL family TDP is 700 W. The RTX 3090 family TDP is 350 W. These values apply to all variants in each family.
What memory bandwidth do the families provide?▾
The H200 NVL family provides 4800 GB/s bandwidth. The RTX 3090 family provides 936 GB/s bandwidth. Both numbers are shared across variants.
Do both variants support NVLink?▾
The H200 NVL variant uses an NVLink bridge. The RTX 3090 variant uses NVLink. Both families list NVLink among their interconnect options.
Which is cheaper to rent, the H200 NVL or the RTX 3090?▾
Cloud rental prices for both the H200 NVL and RTX 3090 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 H200 NVL have compared to the RTX 3090?▾
The H200 NVL has 141 GB of HBM3e memory. The RTX 3090 has 24 GB of GDDR6X memory.
Can I find H200 NVL and RTX 3090 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 H200 NVL and the RTX 3090?▾
The H200 NVL uses the Hopper architecture (2024) while the RTX 3090 uses Ampere (2020). The H200 NVL delivers 13.9x the dense FP16 throughput (989 vs 71 TFLOPS, both without sparsity) and 5.1x the memory bandwidth of the RTX 3090.
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 H200 NVL and the RTX 3090. 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 H200Every current offer by provider, daily price history and a price alert.
- Rent RTX 3090Every 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.