ACS supports multiple GPU families for AI and HPC workloads. Use the alibabacloud.com/gpu-model-series label to request a specific GPU family in your cluster. The following table summarizes the available families.
Choose a GPU family
| Family | GPU memory | Key specs | Best for |
|---|---|---|---|
| GU8TF | 96 GB | FP8, NVLink, 1.6 Tbps RDMA | Large model inference (70B+), small to medium training |
| GU8TEF | 141 GB | FP8, NVLink, 1.6 Tbps RDMA | Large model inference (DeepSeek-LLM 67B), small to medium training |
| L20 (GN8IS) | 48 GB | FP8, TensorRT, P2P | Broad AI workloads, 70B+ inference |
| L20X (GX8SF) | 141 GB | NVLink, 3.2 Tbps RDMA | Large-scale model training and inference |
| P16EN | 96 GB | FP16, 700 GB/s interconnect, 1.6 Tbps RDMA | Large model inference (DeepSeek R1), small to medium training |
| G49E | 48 GB | RTX, TensorRT, P2P | AI and graphics workloads |
| T4 | 16 GB | FP16/INT8/INT4 Tensor Cores, 320 GB/s | Inference and graphics |
| A10 | 24 GB | RTX, TensorRT, Ampere | Deep learning, HPC, graphics |
| G59 | 32 GB | RTX, TensorRT, P2P | AI and HPC workloads |
GU8TF
Compute GPUs with 96 GB of memory per GPU and native FP8 support, enabling single-node inference for 70B and larger models. All 8 GPUs are interconnected via NVLink for small to medium-scale model training. Inter-node communication runs at 1.6 Tbps via Remote Direct Memory Access (RDMA).
Pod resource constraints:
GPU count (memory per GPU) | vCPU | Memory options (GiB) | Memory increment (GiB) | Storage range (GiB) |
1 (96 GB) | 2 | 2–16 | 1 | 30–256 |
4 | 4–32 | 1 | ||
6 | 6–48 | 1 | ||
8 | 8–64 | 1 | ||
10 | 10–80 | 1 | ||
12 | 12–96 | 1 | ||
14 | 14–112 | 1 | ||
16 | 16–128 | 1 | ||
22 | 22, 32, 64, 128 | N/A | ||
2 (96 GB) | 16 | 16–128 | 1 | 30–512 |
32 | 32, 64, 128, 230 | N/A | ||
46 | 64, 128, 230 | N/A | ||
4 (96 GB) | 32 | 32, 64, 128, 256 | N/A | 30–1,024 |
64 | 64, 128, 256, 460 | N/A | ||
92 | 128, 256, 460 | N/A | ||
8 (96 GB) | 64 | 64, 128, 256, 512 | N/A | 30–2,048 |
128 | 128, 256, 512, 920 | N/A | ||
184 | 256, 512, 920 | N/A |
GU8TEF
Compute GPUs with 141 GB of memory per GPU and native FP8 support. Multi-GPU configurations support single-node inference for models such as DeepSeek-LLM 67B. All 8 GPUs are interconnected via NVLink for small to medium-scale model training. Inter-node communication runs at 1.6 Tbps via RDMA.
Pod resource constraints:
GPU count (memory per GPU) | vCPU | Memory options (GiB) | Memory increment (GiB) | Storage range (GiB) |
1 (141 GB) | 2 | 2–16 | 1 | 30–768 |
4 | 4–32 | 1 | ||
6 | 6–48 | 1 | ||
8 | 8–64 | 1 | ||
10 | 10–80 | 1 | ||
12 | 12–96 | 1 | ||
14 | 14–112 | 1 | ||
16 | 16–128 | 1 | ||
22 | 22, 32, 64, 128, 225 | N/A | ||
2 (141 GB) | 16 | 16–128 | 1 | 30–1,536 |
32 | 32, 64, 128, 256 | N/A | ||
46 | 64, 128, 256, 450 | N/A | ||
4 (141 GB) | 32 | 32, 64, 128, 256 | N/A | 30–3,072 |
64 | 64, 128, 256, 512 | N/A | ||
92 | 128, 256, 512, 900 | N/A | ||
8 (141 GB) | 64 | 64, 128, 256, 512 | N/A | 30–6,144 |
128 | 128, 256, 512, 1,024 | N/A | ||
184 | 256, 512, 1024, 1,800 | N/A |
L20 (GN8IS)
Compute GPUs with 48 GB of memory per GPU, suited for a wide range of AI workloads. Supports acceleration libraries such as TensorRT and the FP8 floating-point format. Peer-to-peer (P2P) communication between GPUs is enabled. Multi-GPU configurations support single-node inference for 70B and larger models.
Pod resource constraints:
GPU count (memory per GPU) | vCPU | Memory options (GiB) | Memory increment (GiB) | Storage range (GiB) |
1 (48 GB) | 2 | 2–16 | 1 | 30–256 |
4 | 4–32 | 1 | ||
6 | 6–48 | 1 | ||
8 | 8–64 | 1 | ||
10 | 10–80 | 1 | ||
12 | 12–96 | 1 | ||
14 | 14–112 | 1 | ||
16 | 16–120 | 1 | ||
2 (48 GB) | 16 | 16–128 | 1 | 30–512 |
32 | 32, 64, 128, 230 | N/A | ||
4 (48 GB) | 32 | 32, 64, 128, 256 | N/A | 30–1,024 |
64 | 64, 128, 256, 460 | N/A | ||
8 (48 GB) | 64 | 64, 128, 256, 512 | N/A | 30–2,048 |
128 | 128, 256, 512, 920 | N/A |
L20X (GX8SF)
Compute GPUs with 141 GB of memory per GPU, designed for large-scale model training and inference. All 8 GPUs are interconnected via NVLink. Inter-node communication runs at 3.2 Tbps via RDMA.
Pod resource constraints:
GPU count (memory per GPU) | vCPU | Memory (GiB) | Memory increment (GiB) | Storage range (GiB) |
8 (141 GB) | 184 | 1,800 | N/A | 30–6,144 |
P16EN
Compute GPUs with 96 GB of memory per GPU and FP16 support. Multi-GPU configurations support single-node inference for models such as DeepSeek R1. All 16 GPUs are interconnected at 700 GB/s for small to medium-scale model training. Inter-node communication runs at 1.6 Tbps via RDMA.
Pod resource constraints:
GPU count (memory per GPU) | vCPU | Memory (GiB) | Memory increment (GiB) | Storage range |
1 (96 GB) | 2 | 2–16 | 1 | 30 GB–384 GB |
4 | 4–32 | 1 | ||
6 | 6–48 | 1 | ||
8 | 8-64 | 1 | ||
10 | 10–80 | 1 | ||
2 (96 GB) | 4 | 4–32 | 1 | 30 GB–768 GB |
6 | 6–48 | 1 | ||
8 | 8–64 | 1 | ||
16 | 16–128 | 1 | ||
22 | 32, 64, 128, 225 | N/A | ||
4 (96 GB) | 8 | 8–64 | 1 | 30 GB–1.5 TB |
16 | 16–128 | 1 | ||
32 | 32, 64, 128, 256 | N/A | ||
46 | 64, 128, 256, 450 | N/A | ||
8 (96 GB) | 16 | 16–128 | 1 | 30 GB–3 TB |
32 | 32, 64, 128, 256 | N/A | ||
64 | 64, 128, 256, 512 | N/A | ||
92 | 128, 256, 512, 900 | N/A | ||
16 (96 GB) | 32 | 32, 64, 128, 256 | N/A | 30 GB–6 TB |
64 | 64, 128, 256, 512 | N/A | ||
128 | 128, 256, 512, 1024 | N/A | ||
184 | 256, 512, 1,024, 1,800 | N/A |
G49E
Compute GPUs with 48 GB of memory per GPU, suited for AI and graphics workloads. Supports acceleration libraries such as RTX and TensorRT. P2P communication between GPUs is enabled.
Pod resource constraints:
GPU count (memory per GPU) | vCPU | Memory options (GiB) | Memory increment (GiB) | Storage range (GiB) |
1 (48 GB) | 2 | 2–16 | 1 | 30–256 |
4 | 4–32 | 1 | ||
6 | 6–48 | 1 | ||
8 | 8–64 | 1 | ||
10 | 10–80 | 1 | ||
12 | 12–96 | 1 | ||
14 | 14–112 | 1 | ||
16 | 16–120 | 1 | ||
2 (48 GB) | 16 | 16–128 | 1 | 30–512 |
32 | 32, 64, 128, 230 | N/A | ||
4 (48 GB) | 32 | 32, 64, 128, 256 | N/A | 30–1024 |
64 | 64, 128, 256, 460 | N/A | ||
8 (48 GB) | 64 | 64, 128, 256, 512 | N/A | 30–2048 |
128 | 128, 256, 512, 920 | N/A |
T4
GPUs based on the Turing architecture, with 16 GB of memory per GPU and 320 GB/s of memory bandwidth. Variable-precision Tensor Cores deliver 65 TFLOPS (FP16), 130 TOPS (INT8), and 260 TOPS (INT4), suited for inference and graphics workloads.
Pod resource constraints (single-node family):
GPU count (memory per GPU) | vCPU | Memory options (GiB) | Memory increment (GiB) | Storage range (GiB) |
1 (16 GB) | 2 | 2–8 | 1 | 30–1536 |
4 | 4–16 | 1 | ||
6 | 6–24 | 1 | ||
8 | 8–32 | 1 | ||
10 | 10–40 | 1 | ||
12 | 12–48 | 1 | ||
14 | 14–56 | 1 | ||
16 | 16–64 | 1 | ||
24 | 24, 48, 90 | N/A | ||
2 (16 GB) | 16 | 16–64 | 1 | |
24 | 24, 48, 96 | N/A | ||
32 | 32, 64, 128 | N/A | ||
48 | 48, 96, 180 | N/A |
A10
GPUs based on the Ampere architecture, with 24 GB of memory per GPU. Supports RTX and TensorRT, suited for deep learning, high-performance computing (HPC), and graphics workloads.
Pod resource constraints:
GPU count (memory per GPU) | vCPU | Memory options (GiB) | Memory increment (GiB) | Storage range (GiB) |
1 (24 GB) | 2 | 2–8 | 1 | 30–256 |
4 | 4–16 | 1 | ||
6 | 6–24 | 1 | ||
8 | 8–32 | 1 | ||
10 | 10–40 | 1 | ||
12 | 12–48 | 1 | ||
14 | 14–56 | 1 | ||
16 | 16–60 | 1 | ||
2 (24 GB) | 16 | 16–64 | 1 | 30–512 |
32 | 32, 64, 120 | N/A | ||
4 (24 GB) | 32 | 32, 64, 128 | N/A | 30–1,024 |
64 | 64, 128, 240 | N/A | ||
8 (24 GB) | 64 | 64, 128, 256 | N/A | 30–2,048 |
128 | 128, 256, 480 | N/A |
G59
Compute GPUs with 32 GB of memory per GPU, suited for AI and HPC workloads. Supports RTX and TensorRT. P2P communication between GPUs is enabled. Network bandwidth scales from 1 Gbps per vCPU at smaller configurations to 100 Gbps per vCPU at the largest.
Pod resource constraints:
GPU (Cards) | vCPU | Memory (GiB) | Memory Step Size (GiB) | Storage (GiB) | Networking |
1 (32 GiB GPU memory) | 2 | 2 ~ 16 | 1 | 30 ~ 256 | 1 Gbps per vCPU |
4 | 4 ~ 32 | 1 | |||
6 | 6 ~ 48 | 1 | |||
8 | 8 ~ 64 | 1 | |||
10 | 10 ~ 80 | 1 | |||
12 | 12 ~ 96 | 1 | |||
14 | 14 ~ 112 | 1 | |||
16 | 16 ~ 128 | 1 | |||
22 | 22, 32, 64, 128 | N/A | |||
2 (32 GiB × 2 GPU memory) | 16 | 16 ~ 128 | 1 | 30 ~ 512 | |
32 | 32, 64, 128, 256 | N/A | |||
46 | 64, 128, 256, 360 | N/A | |||
4 (32 GiB × 4 GPU memory) | 32 | 32, 64, 128, 256 | N/A | 30 ~ 1024 | |
64 | 64, 128, 256, 512 | N/A | |||
92 | 128, 256, 512, 720 | N/A | |||
8 (32 GiB × 8 GPU memory) | 64 | 64, 128, 256, 512 | N/A | 30 ~ 2048 | |
128 | 128, 256, 512, 1024 | N/A | 100 Gbps | ||
184 | 256, 512, 1024, 1440 | N/A |