Lambda

Lambda offers cloud-based GPU clusters and superclusters optimized for AI and HPC workloads, featuring high-density, liquid-cooled NVIDIA GPUs. Their offerings focus on flexibility, performance, and ease of use for ML researchers and teams needing scalable compute resources.

Lambda

Provider Profile

Founded

Data Not Available

Headquarters

San Francisco, CA, USA

Pricing Model

$4.49 per GPU per hour, with shared file system storage at $0.20 per GB per month. No additional costs for egress or ingress.

Technical Specification

Target Audience
  • Enterprise ML
  • Researchers
GPU Clusters & Offerings
  • 1-Click Clusters with 16 to 512 NVIDIA H100 Tensor Core GPUs
  • Superclusters featuring from 4,000 to over 165,000 NVIDIA GPUs including the GB300 NVL72 and HGX B300 models
Network Fabric
  • InfiniBand
  • Ethernet
Connectivity Bandwidth
  • 400Gbps
  • NVLink
  • and up to 3200 Gb/s node-to-node compute between worker nodes
Storage Architecture
  • NVMe
  • Shared file system
Compute Framework Compatibility
  • PyTorch
  • TensorFlow
  • CUDA
  • cuDNN
Resource Orchestration
  • Managed Kubernetes
  • Managed Slurm
Support Operations
24/7 Dedicated support
Resource Availability
GA
Datacenter Locations
Key Platform Features
  • NVIDIA Quantum-2 InfiniBand 400 Gb/s
  • High-speed NVLink and NVSwitch interconnections
  • Shared file system storage
  • Shared and redundant 100Gb/s DIA ISP connections
  • Support for NVIDIA's latest GPU architectures

Last Audit: February 2026