Specifications Compared
| Spec | RTX 4090 | H100 |
|---|---|---|
| TDP | 450W | 700W |
| VRAM | 24 GB | 80-94 GB |
| CUDA Cores | 16,384 | 16,896 |
| FP8 (dense) | 330.3 TFLOPS | 1,979 TFLOPS |
| Memory Type | GDDR6X | HBM3 |
| Architecture | Ada Lovelace | Hopper |
| FP16 (dense) | 165.2 TFLOPS | 989 TFLOPS |
| Form Factors | PCIe | SXM5, PCIe, NVL |
| INT8 (dense) | 660.6 TOPS | 1,979 TOPS |
| Interconnect | PCIe 4.0 | NVLink, PCIe 5.0, InfiniBand |
| Tensor Cores | 512 | 528 |
| FP32 Performance | 82.6 TFLOPS | 67 TFLOPS |
| FP64 Performance | 1.3 TFLOPS | 34 TFLOPS |
| Memory Bandwidth | 1,008 GB/s | 3,350 GB/s |
| FP8 (with sparsity) | 660.6 TFLOPS | 3,958 TFLOPS |
| FP16 (with sparsity) | 330.4 TFLOPS | 1,979 TFLOPS |
| INT8 (with sparsity) | 1,321.2 TOPS | 3,958 TOPS |
Performance Analysis
The H100 provides 989 TFLOPS FP16 dense compared with 165.2 TFLOPS FP16 dense on the RTX 4090. This gap widens to 1,979 TFLOPS FP16 with sparsity on the H100 versus 330.4 TFLOPS FP16 with sparsity on the RTX 4090. The RTX 4090 FP32 rating of 82.6 TFLOPS exceeds the H100 FP32 rating of 67 TFLOPS. Higher memory bandwidth of 3350 GB/s on the H100 supports larger batch sizes than the 1008 GB/s bandwidth on the RTX 4090 during training and inference workloads that rely on dense FP16 or sparse FP16 operations.
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.
RTX 4090
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| Vast.ai | United Kingdom, GB | 2 | $0.43 | $0.85 | Deploy |
| RunPod | global | 1 | $0.74 | — | Deploy |
| LeaderGPU | The Netherlands | 8 | $0.88 | $7.04 | Deploy |
3 providers in stock, 8 offers (cheapest per provider shown). All RTX 4090 offers, price history and alerts
H100
| Provider | Region | GPUs | Per GPU / hr | Instance / hr | Deploy |
|---|---|---|---|---|---|
| Vast.ai | Czechia, CZ | 1 | $2.27 | — | Deploy |
| Hyperstack | CANADA-1 | 1 | $2.50 | — | 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 |
9 providers in stock, 24 offers (cheapest per provider shown). All H100 offers, price history and alerts
Notify me when RTX 4090 drops below a price
One email when the cheapest in-stock on-demand price per GPU-hour falls below your threshold. Cheapest right now: $0.43/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 RTX 4090
The RTX 4090 suits workloads that fit within 24 GB memory and run on PCIe form factor systems. Its 82.6 TFLOPS FP32 rating exceeds the H100 FP32 rating and its 450W TDP allows operation in standard consumer power envelopes for tasks such as Stable Diffusion.
When to Choose the H100
The H100 suits workloads that require 80-94 GB memory and 3350 GB/s bandwidth. Its 989 TFLOPS FP16 dense rating and 1,979 TFLOPS FP16 with sparsity rating enable larger models than the corresponding 165.2 TFLOPS and 330.4 TFLOPS figures on the RTX 4090.
Use Cases
The H100 supplies 80-94 GB memory and 989 TFLOPS FP16 dense that exceed the 24 GB and 165.2 TFLOPS FP16 dense of the RTX 4090.
The H100 supplies 3350 GB/s bandwidth and 1,979 TFLOPS FP16 with sparsity that support larger batches than the 1008 GB/s and 330.4 TFLOPS FP16 with sparsity of the RTX 4090.
The H100 supplies 80-94 GB memory that accommodates fine-tuning runs beyond the 24 GB capacity of the RTX 4090.
The RTX 4090 supplies 82.6 TFLOPS FP32 and 450W TDP that fit consumer Stable Diffusion workloads within 24 GB memory.
Frequently Asked Questions
What is the FP16 dense performance difference?▾
The H100 reaches 989 TFLOPS FP16 dense. The RTX 4090 reaches 165.2 TFLOPS FP16 dense.
Which GPU has higher memory bandwidth?▾
The H100 has 3350 GB/s memory bandwidth. The RTX 4090 has 1008 GB/s memory bandwidth.
What are the TDP ratings?▾
The RTX 4090 has a 450W TDP. The H100 has a 700W TDP.
Can the RTX 4090 run models larger than 24 GB?▾
No. The RTX 4090 is limited to 24 GB GDDR6X memory while the H100 provides 80-94 GB HBM3 memory.
Which is cheaper to rent, the RTX 4090 or the H100?▾
Cloud rental prices for both the RTX 4090 and H100 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 RTX 4090 have compared to the H100?▾
The RTX 4090 has 24 GB of GDDR6X memory. The H100 has 80 to 94 GB of HBM3 memory.
Can I find RTX 4090 and H100 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 RTX 4090 and the H100?▾
The RTX 4090 uses the Ada Lovelace architecture (2022) while the H100 uses Hopper (2022). The H100 delivers 6.0x the dense FP16 throughput (989 vs 165.2 TFLOPS, both without sparsity) and 3.3x the memory bandwidth of the RTX 4090.
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 RTX 4090 and the H100. 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 RTX 4090Every current offer by provider, daily price history and a price alert.
- Rent H100Every 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.