lambda-labs-gpu-cloud

1# Lambda Labs GPU Cloud
2 
3Comprehensive guide to running ML workloads on Lambda Labs GPU cloud with on-demand instances and 1-Click Clusters.
4 
5## When to use Lambda Labs
6 
7**Use Lambda Labs when:**
8- Need dedicated GPU instances with full SSH access
9- Running long training jobs (hours to days)
10- Want simple pricing with no egress fees
11- Need persistent storage across sessions
12- Require high-performance multi-node clusters (16-512 GPUs)
13- Want pre-installed ML stack (Lambda Stack with PyTorch, CUDA, NCCL)
14 
15**Key features:**
16- **GPU variety**: B200, H100, GH200, A100, A10, A6000, V100
17- **Lambda Stack**: Pre-installed PyTorch, TensorFlow, CUDA, cuDNN, NCCL
18- **Persistent filesystems**: Keep data across instance restarts
19- **1-Click Clusters**: 16-512 GPU Slurm clusters with InfiniBand
20- **Simple pricing**: Pay-per-minute, no egress fees
21- **Global regions**: 12+ regions worldwide
22 
23**Use alternatives instead:**
24- **Modal**: For serverless, auto-scaling workloads
25- **SkyPilot**: For multi-cloud orchestration and cost optimization
26- **RunPod**: For cheaper spot instances and serverless endpoints
27- **Vast.ai**: For GPU marketplace with lowest prices
28 
29## Quick start
30 
31### Account setup
32 
331. Create account at https://lambda.ai
342. Add payment method
353. Generate API key from dashboard
364. Add SSH key (required before launching instances)
37 
38### Launch via console
39 
401. Go to https://cloud.lambda.ai/instances
412. Click "Launch instance"
423. Select GPU type and region
434. Choose SSH key
44…