USA
1 config
4x
Nvidia GB300
Grace CPU + Blackwell Ultra GPU superchip built for rack-scale AI reasoning and inference.
Compare vs other GPUs →Weekly median price per GPU per hour. Nvidia GB300 price history (JSON).
At a glance
Hardware
- Architecture
- Grace Blackwell Ultra
- Memory per GPU
- 288 GB HBM3e
- Memory bandwidth
- 8,000 GB/s
- Release date
- Q1 2025
Market Updated 15 minutes ago
- Cheapest
- $18.00 / GPU / hr
- on-demand · not verified in stock
- Coverage
- 6 providers
- 26 configs · none verified in stock
GB300 Pricing and Availability
Netherlands
1 config
1x
USA
1 config
72x
Luxembourg
1 config
72x
USA
1 config
72x
Verda
Out of stock
Finland
21 configs
1x-4x
No providers match .
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. Verify before provisioning. More on how we price.
What size AI models can the GB300 run?
One GB300 has 288 GB of VRAM. In practice, that's enough memory for roughly 460B parameters at 4-bit or 125B at 16-bit, assuming a 32K context. Below are some open-weight LLMs, with the estimated memory and GPUs each one needs.
| Model | Parameters | 4-bitINT4 / FP4 | 8-bitFP8 / INT8 | 16-bitFP16 / BF16 |
|---|---|---|---|---|
|
|
27B
|
1 GPU
18 GB VRAM
|
1 GPU
30 GB VRAM
|
1 GPU
58 GB VRAM
|
|
|
30.7B
|
1 GPU
22 GB VRAM
|
1 GPU
35 GB VRAM
|
1 GPU
69 GB VRAM
|
|
|
117B
5.1B active
|
1 GPU
67 GB VRAM
|
1 GPU
120 GB VRAM
|
1 GPU
237 GB VRAM
|
|
|
284B
13B active
|
1 GPU
160 GB VRAM
|
4 GPUs
287 GB VRAM
|
4 GPUs
573 GB VRAM
|
|
|
320B
18B active
|
1 GPU
178 GB VRAM
|
4 GPUs
322 GB VRAM
|
4 GPUs
642 GB VRAM
|
|
|
428B
23B active
|
1 GPU
239 GB VRAM
|
4 GPUs
432 GB VRAM
|
4 GPUs
862 GB VRAM
|
|
|
2.8T
104B active
|
6 GPUs
1,544 GB VRAM
|
11 GPUs
2,804 GB VRAM
|
22 GPUs
5,605 GB VRAM
|
Estimates based on the median on-demand rate. Memory is weights plus FP8 KV cache at 32K context per request (FP16/BF16 in the 16-bit column). GPU counts assume 90% of advertised VRAM is usable. No guarantee of runtime support, usable performance, or that a matching quantized build exists. How we estimate costs.
Technical Specifications
Nvidia GB300 · Per GPU
| Compute · dense | |
|---|---|
| FP4 | 15,000 TFLOPS |
| FP8 | 5,000 TFLOPS 10,000 with sparsity |
| FP16 / BF16 | 2,500 TFLOPS 5,000 with sparsity |
| INT8 | 165 TOPS 330 with sparsity |
| FP32 | 80 TFLOPS |
| FP64 | 1.3 TFLOPS |
| Precision support | FP4FP6FP8FP16BF16TF32FP32FP64INT8 |
| Memory | |
|---|---|
| Capacity | 288 GB HBM3e |
| Bandwidth | 8,000 GB/s |
| Bus width | 8,192-bit |
| ECC | Yes |
| Silicon | |
|---|---|
| Architecture | Blackwell Ultra |
| Process | TSMC 4NP |
| Transistors | 208 billion |
| Shader cores | 20,480 CUDA cores |
| Matrix cores | 640 Tensor cores |
| Compute units | 160 SMs |
| Fabric and host | |
|---|---|
| GPU interconnect | NVLink 1,800 GB/s |
| Host interface | PCIe 6.0 x16 |
| Power | |
|---|---|
| Board power | 1,400 W |
| Cooling | Liquid |
| Platform | |
|---|---|
| Partitioning | MIG, up to 7 instances |
Source: official Nvidia GB300 datasheet.
Frequently Asked Questions
How much does the GB300 cost per hour?
-
As of September 1, 2026, we track 26 configs from 6 providers. Prices are per GPU per hour.
Billing type Configs Median / GPU / hr Cheapest in stock On-demand 4 $18.00 Reserved 16 On request Spot 3 Sold out Custom contract 3 On request No median for on-demand: we only show this when at least 3 providers list the GPU on that billing type. Custom-contract pricing is negotiated per deal and usually not published.
Who has the cheapest GB300?
-
As of September 1, 2026, the lowest listed on-demand price for the GB300 is $18.00 per GPU per hour from Oracle Cloud, though we haven't verified current stock.
Where can I rent a GB300?
-
GetDeploying currently tracks GB300 configs from 6 providers. See the full price comparison above for every provider and config.
How many GB300 GPUs do I need?
-
One GB300 runs models up to roughly 460B parameters at 4-bit quantization or 125B at 16-bit, assuming a 32K context. Larger models run across multiple GPUs: the model table above shows the estimated memory and GPU count for popular open-weight LLMs.
Why choose the GB300?
-
Pairs Grace CPUs with Blackwell Ultra GPUs carrying 288GB HBM3e each, connected in a 72-GPU NVLink domain. Suited for frontier-scale training and long-context reasoning inference where per-GPU memory and interconnect bandwidth are the bottleneck.
Alternatives to Nvidia GB300
Last updated