AMD MI325X
vs
Nvidia A30
Two data center GPUs, one CDNA 3 and one Ampere: 256 GB HBM3e against 24 GB HBM2, 6.4x the memory bandwidth on the MI325X, and Infinity Fabric 896 GB/s on the MI325X where the A30 is PCIe only.
At a glance
Updated 6 minutes ago- Memory
- 256 GB
- HBM3e
- Memory bandwidth
- 6.00 TB/s
- From
- $2.25 /GPU/hr
- 5 providers
- Memory
- 24 GB
- HBM2
- Memory bandwidth
- 933 GB/s
- From
- $0.70 /GPU/hr
- 7 providers
Price comparison
Cheapest listed price per GPU per hour by billing type, and the median across providers.
| Billing | MI325X | A30 | Difference |
|---|---|---|---|
| On-demand | $2.25 | $0.70 | 3.2x more expensive on MI325X |
| Reserved | $1.48 | $0.43 | 3.4x more expensive on MI325X |
| Median | $3.07 | $0.77 | 4x more expensive on MI325X |
Median: Median across providers, one vote each, so a large catalog does not outweigh a small one.
Price history
MI325X vs A30 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | MI325X | A30 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 256 GB HBM3e | 24 GB HBM2 | 11x more on MI325X |
| Bandwidth | 6,000 GB/s | 933 GB/s | 6.4x higher on MI325X |
| Bus width | 8,192-bit | ||
| Compute · dense | |||
| FP8 | 2,614.9 TFLOPS | Not supported | MI325X only |
| FP16 / BF16 | 1,307.4 TFLOPS | 165 TFLOPS | 7.9x higher on MI325X |
| INT8 | 2,614.9 TOPS | 330 TOPS | 7.9x higher on MI325X |
| FP32 | 163.4 TFLOPS | 10.3 TFLOPS | 16x higher on MI325X |
| FP64 | 81.7 TFLOPS | 5.2 TFLOPS | 16x higher on MI325X |
| Platform | |||
| Interconnect | Infinity Fabric 896 GB/s | PCIe only | Infinity Fabric on MI325X only |
| Board power | 1,000 W | 165 W | 6.1x higher on MI325X |
| Released | Q4 2024 | Q2 2021 | 4 years newer on MI325X |
| Precision | |||
| Supported formats | FP8FP16BF16TF32FP32FP64INT8 | FP16BF16TF32FP32FP64INT4INT8 | |
Compute: Vendor peak figures, dense and per GPU. Not measured throughput.
Source: MI325X datasheet and A30 datasheet.
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | MI325X /GPU/hr | A30 /GPU/hr |
|---|---|---|
Bentaus
|
On-Demand $2.25 | -- |
Cyfuture AI
|
On-Demand $3.02 | -- |
DigitalOcean
|
On-Demand $3.80 | -- |
TensorWave
|
Custom $2.25 | -- |
Vultr
|
Spot $2.00 | -- |
Leafcloud
|
-- | On-Demand $0.70 |
Exoscale
|
-- | On-Demand $0.74 |
Leaseweb
|
-- | Reserved (1mo) $0.43 |
AceCloud
|
-- | Reserved (12mo) $0.67 |
Massed Compute
|
-- | On-Demand $0.35 |
Sesterce
|
-- | On-Demand $0.39 |
EmpirioLabs AI
|
-- | On-Demand $0.55 |
Similar GPUs
Other comparisons in the same class.
Our data for AMD MI325X was last updated on Sept. 7, 2026, and for Nvidia A30 on Sept. 7, 2026.