AMD MI300X
vs
Nvidia RTX 3090
A data center GPU against a consumer card, one CDNA 3 and one Ampere: 192 GB HBM3 against 24 GB GDDR6X and 5.7x the memory bandwidth on the MI300X.
At a glance
Memory
- MI300X
- 192 GB
- RTX 3090
- 24 GB
8x more on MI300X
Memory bandwidth
- MI300X
- 5,300 GB/s
- RTX 3090
- 936 GB/s
5.7x higher on MI300X
Cheapest on-demand
- MI300X
- $2.39 /GPU/hr
- RTX 3090
- $0.10 /GPU/hr
24x more expensive on MI300X
Price comparison
Median price per GPU per hour by billing type, with the cheapest listing under it.
| Billing | MI300X | RTX 3090 | Difference |
|---|---|---|---|
| On-demand | $2.91 /GPU/hr from $2.39 | $0.31 /GPU/hr from $0.10 | 9.4x more expensive on MI300X |
| Reserved | $3.39 /GPU/hr from $1.61 (12 mo) | -- /GPU/hr from $0.06 (1 mo) | |
| Spot | -- /GPU/hr from $1.60 | -- /GPU/hr from $0.09 |
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 spot on the MI300X and reserved and spot on the RTX 3090: 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
MI300X vs RTX 3090 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | MI300X | RTX 3090 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 192 GB HBM3 | 24 GB GDDR6X | 8x more on MI300X |
| Bandwidth | 5,300 GB/s | 936 GB/s | 5.7x higher on MI300X |
| Bus width | 8,192-bit | 384-bit | 21x wider on MI300X |
| Compute · dense | |||
| FP8 | 2,614.9 TFLOPS | Not supported | MI300X only |
| FP16 / BF16 | 1,307.4 TFLOPS | 71.2 TFLOPS | 18x higher on MI300X |
| INT8 | 2,614.9 TOPS | 284.7 TOPS | 9.2x higher on MI300X |
| FP32 | 163.4 TFLOPS | 35.6 TFLOPS | 4.6x higher on MI300X |
| FP64 | 81.7 TFLOPS | Not supported | MI300X only |
| Platform | |||
| Interconnect | Infinity Fabric 896 GB/s | NVLink 112 GB/s | 8x higher on MI300X |
| Board power | 750 W | 350 W | 2.1x higher on MI300X |
| Released | Q4 2023 | Q3 2020 | 3 years newer on MI300X |
| Precision | |||
| Supported formats | FP8FP16BF16TF32FP32FP64INT8 | FP16BF16TF32FP32INT4INT8 | |
Compute: Vendor peak figures, dense and per GPU. Not measured throughput.
Source: MI300X datasheet and RTX 3090 datasheet.
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | MI300X /GPU/hr | RTX 3090 /GPU/hr |
|---|---|---|
Runpod
|
On-Demand $2.39 | Reservation on request |
DigitalOcean
|
On-Demand $2.59 | -- |
Cyfuture AI
|
On-Demand $2.78 | -- |
Hot Aisle
|
On-Demand $2.99 | -- |
Oracle Cloud
|
On-Demand $6.00 | -- |
Azure
|
On-Demand $7.86 | -- |
Cirrascale
|
Reserved (12mo) $3.12 | -- |
Crusoe
|
Reservation on request | -- |
HyperAI
|
-- | On-Demand $0.10 |
Theta EdgeCloud
|
-- | On-Demand $0.13 |
Salad
|
-- | On-Demand $0.17 |
Vast.ai
|
-- | Spot $0.11 |
SwissGPU
|
-- | On-Demand $0.31 |
TensorDock
|
-- | On-Demand $0.31 |
EmpirioLabs AI
|
-- | On-Demand $0.40 |
10 providers list the MI300X and 8 the RTX 3090. Every listing is on the MI300X page and the RTX 3090 page.
Similar GPUs
Other comparisons in the same class.
Our data for AMD MI300X was last updated on Sept. 7, 2026, and for Nvidia RTX 3090 on Sept. 7, 2026.