Resource Specification List
Last Update: 2025-04-21 16:09:02
Overview
The following GPU instance specifications cater to high-performance computing, deep learning, and graphics-intensive workloads. For GPU model and Inference model pricings, please refer to Pricing Page.
Application Scenarios
- Large-scale machine learning training and inference
- High-performance computing (HPC): computational finance, quantum simulation, molecular modeling
- AI/GPU-accelerated workloads: rendering, video processing, genomics
Specification List
1. NVIDIA H100 SXM (80GB HBM3)
CPU: Intel Xeon Platinum 8480C
GPU Memory: 80GB per GPU
Storage: Standard
Location: ap-southeast-1a
Instance Type | vCPU | Memory (GB) | GPU Count | GPU Memory (GB) |
|---|---|---|---|---|
h100.1.cpu.20.mem.194 | 20 | 194 | 1 | 80 |
h100.2.cpu.40.mem.388 | 40 | 388 | 2 | 160 |
h100.3.cpu.60.mem.582 | 60 | 582 | 3 | 240 |
h100.4.cpu.80.mem.776 | 80 | 776 | 4 | 320 |
h100.5.cpu.100.mem.970 | 100 | 970 | 5 | 400 |
h100.8.cpu.160.mem.1552 | 160 | 1552 | 8 | 640 |
2. NVIDIA GeForce RTX 3090 (24GB GDDR6X)
CPU: AMD EPYC 7T83
GPU Memory: 24GB per GPU
Storage: Standard
Location: ap-southeast-1a
Instance Type | vCPU | Memory (GB) | GPU Count | GPU Memory (GB) |
|---|---|---|---|---|
rtx3090.1.cpu.20.mem.600 | 20 | 600 | 1 | 24 |
rtx3090.2.cpu.40.mem.1200 | 40 | 1200 | 2 | 48 |
Key Notes
- Availability: Check with your account manager for real-time updates on instance access.
- Storage: All GPU instances default to Standard storage.