The GB300 is newer, with 3.6x the memory and 2.4x the bandwidth. The H100 is cheaper right now.
At a glance
Memory
3.6x more on GB300
- GB300
- 288 GB
- H100
- 80 GB
Memory bandwidth
2.4x higher on GB300
- GB300
- 8.00 TB/s
- H100
- 3.35 TB/s
Cheapest on-demand
90% cheaper on H100
- GB300
- $18.00 /GPU/hr
- H100
- $1.71 /GPU/hr
Price comparison
Median price per GPU per hour by billing type, with the cheapest listing under it.
| Billing | GB300 | H100 | Difference |
|---|---|---|---|
| On-demand | -- /GPU/hr from $18.00 | $3.50 /GPU/hr from $1.71 | |
| Reserved | On request | $3.14 /GPU/hr from $1.53 (1 mo) | |
| Spot | Sold out | $2.00 /GPU/hr from $0.67 |
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 GB300: 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
GB300 vs H100 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | GB300 | H100 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 288 GB HBM3e | 80 GB HBM3 94 GB on NVL | 3.6x more on GB300 |
| Bandwidth | 8,000 GB/s | 3,350 GB/s 2,000 GB/s on PCIe 3,938 GB/s on NVL | 2.4x higher on GB300 |
| Bus width | 8,192-bit | 5,120-bit | 60% wider on GB300 |
| Compute · dense | |||
| FP4 | 15,000 TFLOPS | Not supported | GB300 only |
| FP8 | 5,000 TFLOPS | 1,979 TFLOPS | 2.5x higher on GB300 |
| FP16 / BF16 | 2,500 TFLOPS | 989.5 TFLOPS | 2.5x higher on GB300 |
| INT8 | 165 TOPS | 1,979 TOPS | 12x higher on H100 |
| FP32 | 80 TFLOPS | 67 TFLOPS | 19% higher on GB300 |
| FP64 | 1.3 TFLOPS | 34 TFLOPS | 26x higher on H100 |
| Platform | |||
| Interconnect | NVLink 1,800 GB/s | NVLink 900 GB/s NVLink 600 GB/s on PCIe, NVL | 2x higher on GB300 |
| Board power | 1,400 W | 700 W | 2x higher on GB300 |
| Released | Q1 2025 | Q3 2022 | 2 years newer on GB300 |
| Precision | |||
| Supported formats | FP4FP6FP8FP16BF16TF32FP32FP64INT8 | 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: GB300 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 | GB300 | H100 |
|---|---|---|
Oracle Cloud
USA
|
$18.00 /GPU/hr On-demand NVL | -- |
AWS
USA
|
On request Reserved | -- |
CoreWeave
USA
|
On request Custom NVL | -- |
Gcore
Luxembourg
|
On request Custom NVL | -- |
Nebius
Netherlands
|
On request Custom NVL | -- |
Together AI
USA
|
On request Custom NVL | -- |
Verda
Finland
|
$10.95 /GPU/hr On-demand SXM | -- |
Vast.ai
USA
|
-- | $1.73 /GPU/hr On-demand SXM |
Lium
Distributed
|
-- | $1.75 /GPU/hr On-demand SXM |
HyperAI
Hong Kong
|
-- | $1.80 /GPU/hr On-demand |
GPU.ai
UAE
|
-- | $2.06 /GPU/hr On-demand NVL |
Hyperstack
UK
|
-- | $2.50 /GPU/hr On-demand PCIe |
Massed Compute
USA
|
-- | $2.73 /GPU/hr On-demand |
Spheron
Singapore
|
-- | $2.75 /GPU/hr On-demand NVL |
Runpod
USA
|
-- | $3.19 /GPU/hr On-demand NVL |
7 providers list the GB300 and 57 the H100. Every listing is on the GB300 page and the H100 page.
Similar GPUs
Other comparisons in the same class.
Our data for Nvidia GB300 was last updated on Sept. 14, 2026, and for Nvidia H100 on Sept. 14, 2026.