GPU comparison
MI450 vs B300: specs and rental prices
AMD Instinct MI450 series (MI455X) against NVIDIA B300: 432 GB against 270 GB of memory. Specs from AMD and NVIDIA, prices as of Oct 29, 2026.
In short
- The MI450 lists 432 GB of GPU memory against 270 GB for the B300: 1.6× higher.
- Memory type is HBM4 on the MI450 and HBM3E on the B300.
- The MI450 lists 23,300 GB/s of memory bandwidth against 7,700 GB/s for the B300: 3.03× higher.
- The MI450 lists 10,100 TFLOPS of FP16 / BF16 tensor throughput against 4,500 TFLOPS for the B300: 2.24× higher.
- Not every figure can be compared: at least one vendor page does not publish: TF32 tensor throughput; Maximum power.
- Only the B300 has a fresh on-demand listing ($7.40 per GPU-hour at Hyperstack); the MI450 has none right now.
This summary is generated from the tables below. Prices were last checked Oct 29, 2026. A listing not re-checked for 7 days is greyed and left out of every lowest-price figure; after 14 days it is removed from the page.
Specifications side by side
| Specification | MI450 MI455X | B300, per GPU (HGX B300) | Difference |
|---|---|---|---|
| GPU memory | 432 GB | 270 GB | MI450 1.6× higher |
| Memory type | HBM4 | HBM3E | Different |
| Memory bandwidth | 23,300 GB/s | 7,700 GB/s | MI450 3.03× higher |
| FP8 tensor throughput | 20,100 TFLOPS (basis not stated) | 9,000 TFLOPS (with sparsity) | Not compared: the vendors state the figures on different bases |
| FP16 / BF16 tensor throughput, with sparsity | 10,100 TFLOPS | 4,500 TFLOPS | MI450 2.24× higher |
| TF32 tensor throughput | Not published | 2,200 TFLOPS (with sparsity) | — |
| FP64 throughput | 5 TFLOPS | 1.2 TFLOPS | MI450 4.17× higher |
| Maximum power | Not published | 1,100 W | — |
| GPU-to-GPU interconnect | 3,600 GB/s | 1,800 GB/s | MI450 2× higher |
Compared versions: MI450 MI455X and B300, per GPU (HGX B300). The figure in the stronger cell is shaded; power is shown without a winner. Two throughput figures are compared only when both vendors state them on the same basis (both with sparsity, or both dense), and “derived” means a total the vendor publishes divided by the GPU count it publishes. Full tables with sources: MI450 specifications · B300 specifications.
Rental prices
| Rental price | MI450 | B300 | Difference |
|---|---|---|---|
| Lowest listed on-demand price | No fresh listing | $7.40 per GPU-hour · Hyperstack · checked Oct 29, 2026 | — |
| Lowest listed spot price | No fresh listing | $0.99 per GPU-hour · Nebius · checked Oct 29, 2026 | — |
| Price per GB of memory | Not available | $0.027 per GB of GPU memory per hour | — |
| Fresh listings | 0 from 0 providers | 13 from 10 providers | — |
Prices cover every version and instance size the providers list, so they are not tied to the versions compared above; the price per GB of memory divides the lowest fresh on-demand price by the memory of the compared version. Listed prices are not like for like across providers. See every listing: MI450 prices · B300 prices.
More on these GPUs
Cite this page
Last updated Oct 29, 2026. Figures may change as records are updated — include the date you retrieved them.
| APA | Data Center Census. (2026). MI450 vs B300. Retrieved from https://datacentercensus.com/gpu-prices/compare/mi450-vs-b300 |
|---|---|
| MLA | “MI450 vs B300.” Data Center Census, Oct 29, 2026, https://datacentercensus.com/gpu-prices/compare/mi450-vs-b300. |
| Link | https://datacentercensus.com/gpu-prices/compare/mi450-vs-b300 |
Republishing a figure or table? Attribute it to Data Center Census with a link to this page. The underlying data is downloadable as JSON and CSV; the press kit has standing figures and contact details.