A consumer GPU against a workstation card, one Ada Lovelace and one Turing: 24 GB GDDR6X against 16 GB GDDR6, 2.3x the memory bandwidth on the RTX 4090, and NVLink 50 GB/s on the RTX 5000 where the RTX 4090 is PCIe only.
At a glance
Memory
- RTX 4090
- 24 GB
- RTX 5000
- 16 GB
1.5x more on RTX 4090
Memory bandwidth
- RTX 4090
- 1,010 GB/s
- RTX 5000
- 448 GB/s
2.3x higher on RTX 4090
Cheapest on-demand
- RTX 4090
- $0.33 /GPU/hr
- RTX 5000
- $0.60 /GPU/hr
45% cheaper on RTX 4090
Price comparison
Median price per GPU per hour by billing type, with the cheapest listing under it.
| Billing | RTX 4090 | RTX 5000 | Difference |
|---|---|---|---|
| On-demand | $0.52 /GPU/hr from $0.33 | -- /GPU/hr from $0.60 | |
| Reserved | $0.52 /GPU/hr from $0.27 (12 mo) | No reserved listings | |
| Spot | $0.17 /GPU/hr from $0.16 | No spot listings |
Heads up: A provider's own page may quote a different figure, for example on tax, region or a promotion. A monthly-only plan shows a derived hourly rate. Medians are across providers, one vote each, so a large catalog does not outweigh a small one. No median for on-demand on the RTX 5000: we only show one when at least 3 providers list the GPU on that billing type. Verify before provisioning. More on how we price.
Price history
RTX 4090 vs RTX 5000 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | RTX 4090 | RTX 5000 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 24 GB GDDR6X | 16 GB GDDR6 | 1.5x more on RTX 4090 |
| Bandwidth | 1,010 GB/s | 448 GB/s | 2.3x higher on RTX 4090 |
| Bus width | 384-bit | 256-bit | 1.5x wider on RTX 4090 |
| Compute · dense | |||
| FP8 | 330.3 TFLOPS | Not supported | RTX 4090 only |
| FP16 / BF16 | 165.2 TFLOPS | 89.2 TFLOPS | 1.9x higher on RTX 4090 |
| INT8 | 660.6 TOPS | 178.4 TOPS | 3.7x higher on RTX 4090 |
| FP32 | 82.6 TFLOPS | 11.2 TFLOPS | 7.4x higher on RTX 4090 |
| Platform | |||
| Interconnect | PCIe only | NVLink 50 GB/s | NVLink on RTX 5000 only |
| Board power | 450 W | 265 W | 1.7x higher on RTX 4090 |
| Released | Q4 2022 | Q3 2018 | 4 years newer on RTX 4090 |
| Precision | |||
| Supported formats | FP8FP16BF16TF32FP32INT4INT8 | FP16FP32INT4INT8 | |
Compute: Vendor peak figures, dense and per GPU. Not measured throughput.
Source: RTX 4090 datasheet and RTX 5000 datasheet.
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | RTX 4090 /GPU/hr | RTX 5000 /GPU/hr |
|---|---|---|
Novita
|
On-Demand $0.33 | -- |
Salad
|
On-Demand $0.33 | -- |
GPU.ai
|
On-Demand $0.34 | -- |
Runpod
|
On-Demand $0.34 | -- |
Packet·ai
|
On-Demand $0.39 | -- |
Lium
|
On-Demand $0.40 | -- |
Vast.ai
|
On-Demand $0.40 | -- |
GPUhub
|
On-Demand $0.44 | -- |
OVH
|
-- | On-Demand $0.60 |
Paperspace
|
-- | On-Demand $0.82 |
19 providers list the RTX 4090 and 2 the RTX 5000. Every listing is on the RTX 4090 page and the RTX 5000 page.
Similar GPUs
Other comparisons in the same class.
-
Nvidia RTX 5090
From $0.35 /GPU/hr Compare -
Nvidia L40S
From $0.38 /GPU/hr Compare -
Nvidia RTX 3090
From $0.10 /GPU/hr Compare -
Nvidia A5000
From $0.25 /GPU/hr Compare -
Nvidia RTX 5000 Ada
From $2.12 /GPU/hr Compare
Our data for Nvidia RTX 4090 was last updated on Sept. 7, 2026, and for Nvidia RTX 5000 on Sept. 7, 2026.