AMD MI450X
vs
Nvidia H100
The MI450X is newer, with 3.6x the memory and 7x the bandwidth.
At a glance
Memory
3.6x more on MI450X
- MI450X
- 288 GB
- H100
- 80 GB
Memory bandwidth
7x higher on MI450X
- MI450X
- 23.30 TB/s
- H100
- 3.35 TB/s
Cheapest on-demand
- MI450X
- --
- H100
- $0.39 /GPU/hr
Price comparison
Median price per GPU per hour by billing type, with the cheapest listing under it.
| Billing | MI450X | H100 | Difference |
|---|---|---|---|
| On-demand | No on-demand listings | $3.87 /GPU/hr from $0.39 | |
| Reserved | No reserved listings | $3.17 /GPU/hr from $1.44 (24 mo) | |
| Spot | No spot listings | $2.43 /GPU/hr from $0.75 |
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. Verify before provisioning. More on how we price.
Price history
MI450X vs H100 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | MI450X | H100 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 288 GB HBM4 | 80 GB HBM3 94 GB on NVL | 3.6x more on MI450X |
| Bandwidth | 23,300 GB/s | 3,350 GB/s 2,000 GB/s on PCIe 3,938 GB/s on NVL | 7x higher on MI450X |
| Bus width | 5,120-bit | ||
| Compute · dense | |||
| FP4 | 40,265 TFLOPS | Not supported | MI450X only |
| FP8 | 20,133 TFLOPS | 1,979 TFLOPS | 10x higher on MI450X |
| FP16 / BF16 | 5,033 TFLOPS | 989.5 TFLOPS | 5.1x higher on MI450X |
| INT8 | 5,033 TOPS | 1,979 TOPS | 2.5x higher on MI450X |
| FP32 | 315 TFLOPS | 67 TFLOPS | 4.7x higher on MI450X |
| FP64 | Not supported | 34 TFLOPS | H100 only |
| Platform | |||
| Interconnect | Infinity Fabric 3,600 GB/s | NVLink 900 GB/s NVLink 600 GB/s on PCIe, NVL | 4x higher on MI450X |
| Released | Q4 2026 | Q3 2022 | 4 years newer on MI450X |
| Precision | |||
| Supported formats | FP4FP6FP8FP16BF16FP32INT8 | FP8FP16BF16TF32FP32FP64INT8 | |
Figures are for each model's headline configuration. Where a form factor carries a different number, it is named under the figure.
Compute: Vendor peak figures, dense and per GPU. Not measured throughput.
Source: MI450X datasheet and H100 datasheet.
Where to rent them
Providers that rent these GPUs, with each one's cheapest rate. No provider we track offers both right now.
| Provider | MI450X | H100 |
|---|---|---|
|
|
-- | $2.20 /GPU/hr On-demand SXM |
|
|
-- | $2.48 /GPU/hr On-demand SXM |
|
|
-- | $2.50 /GPU/hr On-demand PCIe |
|
|
-- | $2.50 /GPU/hr On-demand SXM |
|
|
-- | $2.73 /GPU/hr On-demand |
|
|
-- | $3.00 /GPU/hr On-demand SXM |
|
|
-- | $3.15 /GPU/hr On-demand SXM |
|
|
-- | $3.20 /GPU/hr On-demand SXM |
0 providers list the MI450X and 65 the H100. Every listing is on the MI450X page and the H100 page.
Similar GPUs
Other comparisons in the same class.
Our data for AMD MI450X was last updated on Nov. 17, 2026, and for Nvidia H100 on Sept. 14, 2026.