NVIDIA RTX 4090 GPU: Specs, VRAM, Price & Benchmarks (2026)
The NVIDIA RTX 4090 (24GB GDDR6X) launched in 2022. This guide covers its full specs, VRAM, SXM/PCIe differences where they apply, AI performance, and live per-GPU rental pricing on Aquanode.
Short answer on cost: the RTX 4090 rents from $0.393 per GPU per hour on Aquanode.
How much VRAM does the RTX 4090 have?
The RTX 4090 has 24GB GDDR6X, with 1,008 GB/s of derived memory bandwidth (bus width × transfer rate; not a vendor-stated figure).
- VRAM: RTX 4090 24GB GDDR6X vs RTX 5090 32GB GDDR7
- VRAM: RTX 4090 24GB GDDR6X vs RTX 4080 16GB GDDR6X
What fits in 24GB GDDR6X of VRAM
| Model | Precision | Fits? |
|---|---|---|
| Llama 3 8B | FP16 | ~16GB. Fits with a moderate KV-cache budget. |
| Llama 3 8B | INT4 | ~5-6GB. Fits comfortably. |
| Mistral 7B | FP16 | ~14GB. Fits. |
| Llama 3.1 70B | INT4 | ~35-40GB. Does not fit on a single 24GB card. |
Approximate, based on published parameter counts and standard bytes-per-parameter rules of thumb (FP16 ≈ 2 bytes/param, INT4 ≈ 0.5-0.6 bytes/param). Real footprint also depends on KV-cache size and framework overhead.
What can the RTX 4090 run?
Popular open models from small to frontier scale, with the memory each needs and how many RTX 4090 cards (24GB GDDR6X each) that takes.
| Model | As published | FP8 | INT4 |
|---|---|---|---|
| Qwen/Qwen3-8B 8.2B | BF16: ~18.3 GB, 1 GPU | FP8: ~9.2 GB, 1 GPU | INT4: ~4.6 GB, 1 GPU |
| Qwen/Qwen2.5-14B-Instruct 14.8B | BF16: ~33 GB, 2 GPUs | FP8: ~16.5 GB, 1 GPU | INT4: ~8.3 GB, 1 GPU |
| Qwen/Qwen3-32B 32.8B | BF16: ~73.2 GB, 4 GPUs | FP8: ~36.6 GB, 2 GPUs | INT4: ~18.3 GB, 1 GPU |
| Qwen/Qwen-72B 72.3B | BF16: ~162 GB, 7 GPUs | FP8: ~80.8 GB, 4 GPUs | INT4: ~40.4 GB, 2 GPUs |
| MiniMaxAI/MiniMax-M2.7 228.7B | FP8: ~256 GB, 11 GPUs | – | INT4: ~128 GB, 6 GPUs |
| deepseek-ai/DeepSeek-R1 684.5B | FP8: ~765 GB, 32 GPUs | – | INT4: ~383 GB, 16 GPUs |
Estimates: weights at the stated precision plus a flat 20% for KV cache and overhead, at a moderate context length. A dash means the precision is not offered for that model (it is already published at that size). INT4 needs a published quantized checkpoint. Open any model for a per-GPU breakdown, or use the RTX 4090 VRAM calculator.
RTX 4090 VRAM calculator: check which models fit in its memory at each precision.
All models that fit in 24 GB: the open models whose weights and overhead fit, at native, FP8 and INT4 precision.
RTX 4090 specs
| Architecture | NVIDIA, launched 2022 |
| VRAM | 24GB GDDR6X |
| Memory bandwidth | 1,008 GB/s (derived) |
| TDP | 450W |
| Form factor | PCIe, 3-slot |
Specs from the vendor's public product page. Memory bandwidth is marked (derived) because NVIDIA does not publish a bandwidth figure for this part. It is bus width × transfer rate, not a vendor-stated number. See the source.
Related reading: RTX 4090 for AI, RTX 5090 vs RTX 4090, and The best GPUs for AI, ranked.
GPU Glossary: What is VRAM?, Tensor Cores, CUDA Cores, TFLOPS
RTX 4090 AI performance
A consumer flagship that's genuinely useful for renting by the hour: strong FP16/INT8 throughput and 1,008 GB/s bandwidth make it capable for fine-tuning and inference on models that fit in 24GB, at a lower hourly rate than data-center cards.
- Memory bandwidth: 1,008 GB/s (derived)
No ECC memory, no NVLink, and a 24GB ceiling. Not built for multi-GPU training or for any workload where a silent memory bit-flip in a long run is unacceptable. Availability and TDP (450W) also make it a poor fit for dense multi-card racks.
The cheapest RTX 4090 offer right now ($0.393/GPU/hr) is about 43% below the market median of $0.693/GPU/hr.
RTX 4090 price: what does it cost?
Buying. $1,599 (launch MSRP).
Renting. The cheapest current on-demand rate for the RTX 4090 on Aquanode is $0.393/GPU/hr (no on-demand supply right now; this is the cheapest offer overall). Live rates range from $0.393 to $1.22 per GPU per hour, with a median of $0.693/GPU/hr. Billed by the offer's own terms; the table below shows every live rate.
| Region | $/GPU/hr | Available | VRAM | vCPU | RAM |
|---|---|---|---|---|---|
| Iev, Ua | $0.393 | 1 | 24 GB | 35 | 42 GB |
| United States | $0.418 | 1 | 24 GB | 16 | 126 GB |
| California, Us | $0.573 | 8 | 24 GB | 24 | 47.3 GB |
RTX 4090 price history
Aquanode stores one snapshot of its GPU price index per UTC day. For the RTX 4090 that is 25 days so far, 2026-10-06 to 2026-10-30, so this is a short history, not a long-run trend. The lowest per-GPU rate went from $0.374 on 2026-10-06 to $0.418 on 2026-10-30.
| Day (UTC) | Lowest per GPU hour | Median per GPU hour | Data-center lowest per GPU hour | Offers |
|---|---|---|---|---|
| 2026-10-30 | $0.418 | $0.707 | $0.761 | 59 |
| 2026-10-29 | $0.393 | $0.736 | $0.761 | 49 |
| 2026-10-28 | $0.393 | $0.801 | $0.760 | 57 |
| 2026-10-27 | $0.393 | $0.781 | $0.760 | 35 |
| 2026-10-26 | $0.393 | $0.771 | $0.761 | 39 |
| 2026-10-25 | $0.393 | $0.793 | – | 42 |
| 2026-10-24 | $0.495 | $0.957 | – | 15 |
| 2026-10-23 | $0.393 | $1.18 | – | 23 |
| 2026-10-22 | $0.393 | $1.22 | $0.760 | 14 |
| 2026-10-21 | $0.692 | $0.809 | $0.760 | 31 |
| 2026-10-20 | $0.393 | $0.763 | $0.760 | 47 |
| 2026-10-19 | $0.393 | $0.810 | $0.761 | 40 |
| 2026-10-18 | $0.393 | $0.763 | $0.760 | 41 |
| 2026-10-17 | $0.479 | $0.780 | $0.763 | 37 |
| 2026-10-16 | $0.374 | $0.979 | $0.760 | 52 |
| 2026-10-15 | $0.374 | $0.979 | $0.761 | 36 |
| 2026-10-14 | $0.374 | $0.769 | $0.761 | 58 |
| 2026-10-13 | $0.374 | $0.728 | $0.979 | 84 |
| 2026-10-12 | $0.374 | $0.736 | $0.761 | 85 |
| 2026-10-11 | $0.374 | $0.772 | $0.761 | 96 |
| 2026-10-10 | $0.374 | $0.753 | $0.761 | 93 |
| 2026-10-09 | $0.374 | $0.721 | $0.761 | 82 |
| 2026-10-08 | $0.374 | $0.736 | $0.761 | 87 |
| 2026-10-07 | $0.374 | $0.632 | $0.761 | 126 |
| 2026-10-06 | $0.374 | $0.642 | $0.761 | 93 |
Each row is the stored daily snapshot of the live index, copied as recorded. A dash means that day stored no figure. The same series is available as JSON and summarised in the monthly GPU price report.
How this price is calculated
All prices on this page are normalized to a per-GPU hourly rate using each offer's authoritative GPU count, so that raw price is divided by the number of GPUs it actually covers; some offers report price as already per-GPU, so those are used as-listed. An offer with a missing, zero, or invalid GPU count is excluded entirely rather than published at a guessed rate.
No offers were excluded from this snapshot for a missing or invalid price. 1 offer(s) were dropped as high-side price outliers (more than 3x the model's median rate).
Only the cheapest qualifying offer per provider is shown in the table above. This page regenerates at most once per hour.
RTX 4090 vs RTX 5090: how do they compare?
- VRAM: RTX 4090 24GB GDDR6X vs RTX 5090 32GB GDDR7
- Memory bandwidth: 1,008 GB/s vs 1,792 GB/s
On Aquanode right now, RTX 4090 starts at $0.393/GPU/hr against RTX 5090's $0.658/GPU/hr, about 40% less.
RTX 4090 vs RTX 4080: how do they compare?
- VRAM: RTX 4090 24GB GDDR6X vs RTX 4080 16GB GDDR6X
- Memory bandwidth: 1,008 GB/s vs 716.8 GB/s
On Aquanode right now, RTX 4090 starts at $0.393/GPU/hr against RTX 4080's $0.620/GPU/hr, about 37% less.
Compare the RTX 4090 with other GPUs
Compare RTX 4090 with
- B200 vs RTX 4090
- H200 vs RTX 4090
- H100 vs RTX 4090
- A100 vs RTX 4090
- DGX A100 vs RTX 4090
- RTX 4090 vs V100
- AMD MI300X vs RTX 4090
- L40S vs RTX 4090
- L40 vs RTX 4090
- A40 vs RTX 4090
- RTX 4090 vs RTX A6000
- RTX 4090 vs RTX A5000
- RTX 4090 vs RTX A4000
- L4 vs RTX 4090
- RTX 4090 vs T4
- RTX 4090 vs RTX 6000 Ada
- RTX 4090 vs RTX PRO 6000
- RTX 4090 vs RTX PRO 6000 WS
- RTX 4090 vs RTX PRO 6000 SE
- RTX 4090 vs RTX PRO 5000
- RTX 4090 vs RTX 5090
- RTX 4090 vs RTX 5080
- RTX 4090 vs RTX 5070 Ti
- RTX 4090 vs RTX 5070
- RTX 4090 vs RTX 5060 Ti
- RTX 4080 Super vs RTX 4090
- RTX 4080 vs RTX 4090
- RTX 4070 Ti vs RTX 4090
- RTX 4070 Super vs RTX 4090
- RTX 4070 vs RTX 4090
- RTX 4060 Ti vs RTX 4090
- RTX 3090 vs RTX 4090
- RTX 3080 vs RTX 4090
- RTX 3070 vs RTX 4090
- RTX 3060 vs RTX 4090
- B300 vs RTX 4090
- AMD MI355X vs RTX 4090
- AMD MI325X vs RTX 4090
- H100 NVL vs RTX 4090
- RTX 4090 vs RTX PRO 4500
- RTX 4090 vs RTX PRO 4500 SE
- RTX 4090 vs RTX PRO 4000
- RTX 4090 vs RTX 5880 Ada
- RTX 4090 vs RTX 5000 Ada
- RTX 4000 Ada vs RTX 4090
- RTX 4000 SFF Ada vs RTX 4090
- RTX 2000 Ada vs RTX 4090
- RTX 4070 Ti Super vs RTX 4090
- RTX 4090 vs RTX A4500
- RTX 2080 Ti vs RTX 4090
- RTX 3080 Ti vs RTX 4090
- RTX 3070 Ti vs RTX 4090
- RTX 3060 Ti vs RTX 4090
- GTX 1080 Ti vs RTX 4090
- RTX 4060 vs RTX 4090
- RTX 4090 vs RTX 5060
- GTX 1660 Super vs RTX 4090
- RTX 2060 vs RTX 4090
- Radeon RX 9070 XT vs RTX 4090
- Radeon RX 6700 XT vs RTX 4090
- RTX 4090 vs RTX 6000
- A16 vs RTX 4090
- P40 vs RTX 4090
- P4 vs RTX 4090
- Quadro P2000 vs RTX 4090
- Quadro M4000 vs RTX 4090
Get notified when the price drops
GPU supply moves hourly. Tell us what you're waiting for and we'll email you when a matching offer appears across any provider we track.
Rent a RTX 4090 on Aquanode
- On-demand instances from $0.393/GPU/hr, billed by the provider's own terms, with no hardware procurement or long-term commitment.
- 83 live offers across 3 regions today.
- Set a price/availability alert above to hear the moment a cheaper or newly-available RTX 4090 offer appears.
- Compare every RTX 4090 offer side by side, or browse the full multi-provider GPU marketplace.
Good for
A consumer flagship that's genuinely useful for renting by the hour: strong FP16/INT8 throughput and 1,008 GB/s bandwidth make it capable for fine-tuning and inference on models that fit in 24GB, at a lower hourly rate than data-center cards.
Not good for
No ECC memory, no NVLink, and a 24GB ceiling. Not built for multi-GPU training or for any workload where a silent memory bit-flip in a long run is unacceptable. Availability and TDP (450W) also make it a poor fit for dense multi-card racks.
RTX 4090 FAQs
How much VRAM does the RTX 4090 have?
The RTX 4090 has 24GB GDDR6X, with 1,008 GB/s of derived (bus width × transfer rate, not vendor-stated) memory bandwidth.
What is the NVIDIA RTX 4090?
The NVIDIA RTX 4090 is a GPU released in 2022, with 24GB GDDR6X of memory and a 450W power envelope. See the full spec table above for interconnect, form factor and tensor-throughput details.
How much does it cost to rent a RTX 4090?
Live RTX 4090 rental prices currently range from $0.393 to $1.22 per GPU per hour, with a median of $0.693 per GPU per hour.
What's the cheapest RTX 4090 rate?
The lowest current RTX 4090 rate on Aquanode is $0.393 per GPU per hour in Iev, Ua.
How does the RTX 4090 compare to the RTX 5090?
VRAM: RTX 4090 24GB GDDR6X vs RTX 5090 32GB GDDR7 Memory bandwidth: 1,008 GB/s vs 1,792 GB/s On Aquanode right now, RTX 4090 starts at $0.393/GPU/hr against RTX 5090's $0.658/GPU/hr, about 40% less. See the full RTX 4090 vs RTX 5090 comparison for a shared-provider price breakdown.
How does the RTX 4090 compare to the RTX 4080?
VRAM: RTX 4090 24GB GDDR6X vs RTX 4080 16GB GDDR6X Memory bandwidth: 1,008 GB/s vs 716.8 GB/s On Aquanode right now, RTX 4090 starts at $0.393/GPU/hr against RTX 4080's $0.620/GPU/hr, about 37% less. See the full RTX 4090 vs RTX 4080 comparison for a shared-provider price breakdown.
What is the RTX 4090 good for?
A consumer flagship that's genuinely useful for renting by the hour: strong FP16/INT8 throughput and 1,008 GB/s bandwidth make it capable for fine-tuning and inference on models that fit in 24GB, at a lower hourly rate than data-center cards.
What are the RTX 4090's limitations?
No ECC memory, no NVLink, and a 24GB ceiling. Not built for multi-GPU training or for any workload where a silent memory bit-flip in a long run is unacceptable. Availability and TDP (450W) also make it a poor fit for dense multi-card racks.
Is renting cheaper than buying?
Renting avoids the upfront hardware cost and lets you match spend to actual usage. A rented RTX 4090 at $0.393/hr only costs money while it's running, whereas buying ties up capital in hardware that keeps depreciating whether it's in use or not. Buying outright runs $1,599 (launch MSRP). Which is cheaper depends on how continuously you'd run it; short or bursty workloads usually favor renting.
How is the RTX 4090 price calculated?
All prices are normalized to a per-GPU hourly rate using each offer's authoritative GPU count, which the raw price is divided by; some offers report price as already per-GPU. Offers whose price can't be safely normalized, or whose rate is an extreme outlier against the rest of the market, are excluded.
RTX 4090 price by region
Related guides
Other models in the same generation, then the rest of the GPU index.
