Lambda

Lambda announced a multiyear, multibillion-dollar Microsoft infrastructure agreement on November 3, 2025 covering tens of thousands of NVIDIA GPUs including GB300 NVL72 systems.

Lambda offers AI cloud infrastructure from on-demand GPU instances to 16-2,000+ GPU clusters and 4,000-165,000+ GPU superclusters for frontier training and inference.

Lambda offers AI cloud infrastructure from on-demand GPU instances to 16-2,000+ GPU clusters and 4,000-165,000+ GPU superclusters for frontier training and inference.

Lambda reports official MLPerf Training v5.1 submissions on 72 GB300 accelerators, with Llama 3.1 8B time-to-convergence of 14.25 minutes and Llama 2-70B LoRA of 1.26 minutes, and publishes a reproducibility guide.

Lambda is integrating next-generation NVIDIA Vera Rubin, BlueField-4 and photonics technologies into production AI factories for reasoning and agentic workloads, showing strong systems engineering innovation rather than frontier-model research.

Lambda is integrating next-generation NVIDIA Vera Rubin, BlueField-4 and photonics technologies into production AI factories for reasoning and agentic workloads, showing strong systems engineering innovation rather than frontier-model research.

Lambda announced on November 18, 2025 that it had raised over $1.5 billion in Series E equity funding led by TWG Global, with USIT and existing investors participating.

Lambda supports single-tenant superclusters up to 165,000+ GPUs and closed a $926M secured financing facility in August 2026 to fund committed GPU infrastructure deployments.

Lambda supports single-tenant superclusters up to 165,000+ GPUs and closed a $926M secured financing facility in August 2026 to fund committed GPU infrastructure deployments.

Official managed Slurm documentation describes sbatch multi-GPU jobs, Pyxis/Enroot containers and NCCL multi-node all-reduce examples. These standard scheduler, container and collective-communication practices support a distinct inference of portable AI infrastructure skills.

Lambda roles and systems span GPU cloud, distributed training, inference, networking, storage and AI-factory operations; these are highly portable AI infrastructure skills, though more specialized than general software or model research.

Lambda roles and systems span GPU cloud, distributed training, inference, networking, storage and AI-factory operations; these are highly portable AI infrastructure skills, though more specialized than general software or model research.