Billing Description
Cloud GPU Service is pay-as-you-go. See Instance Billing Modes for details.
A GPU instance consists of three major parts: network, storage (system disks and data disks), and computing (CPU, memory, and GPU). The prices shown in the table below are only for computing (CPU, memory, and GPU). For more information on network pricing, please see Public Network Billing. For more information on disk pricing, please see the CBS Pricing List.
Cloud GPU Service is available in the following instance types: Computing GN6S, GN7, GN8, GN10X, and GN10Xp. You can select the GPU instances that best meet your use case according to the specifications and prices. For more information on GPU instance types, please see NVIDIA GPU Instances.
Note：
The prices on this page are for Guangzhou Zone 3. Prices are different depending on the AZ and OS. Please refer to the prices on the purchase page.
Computing GN6S
| Instance Specification |
GPU (Tesla P4) |
GPU Video Memory |
vCPU |
Memory |
Pay-as-You-Go |
| GN6S.LARGE20 |
1 |
8 GB |
4 cores |
20 GB |
1.11 USD/hr |
| GN6S.2XLARGE40 |
2 |
16 GB |
8 cores |
40 GB |
2.22 USD/hr |
Computing GN7
| Instance Specification |
GPU (NVIDIA T4) |
GPU Video Memory |
vCPU |
Memory |
Pay-as-You-Go |
| GN7.LARGE20 |
1/4 vGPU |
4 GB |
4 |
20 |
0.49 USD/hr |
| GN7.2XLARGE40 |
1/2 vGPU |
8 GB |
10 |
40 |
1.01 USD/hr |
| GN7.2XLARGE32 |
1 |
16 GB |
8 |
32 |
1.26 USD/hr |
| GN7.5XLARGE80 |
1 |
16 GB |
20 |
80 |
2.01 USD/hr |
| GN7.8XLARGE128 |
1 |
16 GB |
32 |
128 |
2.77 USD/hr |
| GN7.10XLARGE160 |
2 |
32 GB |
40 |
160 |
4.03 USD/hr |
| GN7.20XLARGE320 |
4 |
64 GB |
80 |
320 |
8.06 USD/hr |
Computing GN8
| Instance Specification |
GPU (Tesla P40) |
GPU Video Memory |
vCPU |
Memory |
Pay-as-You-Go |
| GN8.LARGE56 |
1 |
24 GB |
6 cores |
56 GB |
2.45 USD/hr |
| GN8.3XLARGE112 |
2 |
48 GB |
14 cores |
112 GB |
5.01 USD/hr |
| GN8.7XLARGE224 |
4 |
96 GB |
28 cores |
224 GB |
10.02 USD/hr |
| GN8.14XLARGE448 |
8 |
192 GB |
56 cores |
448 GB |
20.04 USD/hr |
Computing GN10X
| Instance Specification |
GPU (Tesla V100-NVLINK-32G) |
GPU Video Memory |
vCPU |
Memory |
Pay-as-You-Go |
| GN10X.2XLARGE40 |
1 |
32 GB |
8 cores |
40 GB |
2.68 USD/hr |
| GN10X.4XLARGE80 |
2 |
64 GB |
18 cores |
80 GB |
5.43 USD/hr |
| GN10X.9XLARGE160 |
4 |
128 GB |
36 cores |
160 GB |
10.86 USD/hr |
| GN10X.18XLARGE320 |
8 |
256 GB |
72 cores |
320 GB |
21.71 USD/hr |
Computing GN10Xp
| Instance Specification |
GPU (Tesla V100-NVLINK-32G) |
GPU Video Memory |
vCPU |
Memory |
Pay-as-You-Go |
| GN10Xp.2XLARGE40 |
1 |
32 GB |
10 cores |
40 GB |
1.72 USD/hr |
| GN10Xp.5XLARGE80 |
2 |
64 GB |
20 cores |
80 GB |
3.44 USD/hr |
| GN10Xp.10XLARGE160 |
4 |
128 GB |
40 cores |
160 GB |
6.89 USD/hr |
| GN10Xp.20XLARGE320 |
8 |
256 GB |
80 cores |
320 GB |
13.78 USD/hr |
Repossession Description
GPU instances are repossessed in the same manner as CVM instances. For more information, please see CVM Instance Repossession.
Note on Overdue Payment
After your account becomes overdue, GPU instances will be dealt with in the same manner as CVM instances. For more information, please see CVM Overdue Payment.
Notes
Yes
No
Was this page helpful?