# Verda Verda is a full-stack AI cloud for running AI workloads on GPU infrastructure. It provides on-demand NVIDIA GPU instances and self-service multi-node clusters with managed Slurm or Kubernetes. - Findings24: - Site: - Preços: Freemium - Categorias: Nuvem de GPU ## O que é Verda? Verda is a full-stack AI cloud focused on infrastructure for AI workloads. It combines on-demand NVIDIA GPU instances with self-service, multi-node clusters and managed Slurm or Kubernetes options. Users can choose individual GPU capacity for short-lived or focused jobs, or provision a multi-node environment when an application requires distributed compute. ## O que Verda pode fazer? ### On-demand NVIDIA GPU instances Provision single-instance GPU capacity in under a minute, with selectable configurations from 1x to 8x GPUs. ### Multiple NVIDIA GPU options The listed instance hardware includes NVIDIA B300, B200, H200, H100, and A100 GPUs, allowing users to select from the models exposed by the platform. ### Self-service multi-node clusters Create clusters without a manual setup process or a sales call, supporting workloads that need more than one node. ### Managed Slurm or Kubernetes Choose between managed Slurm and Kubernetes for cluster operation, depending on the workload or orchestration workflow. ### Pay-as-you-go GPU usage GPU instances are offered with pay-as-you-go pricing; the available evidence does not state the applicable rates. ## Casos de uso ### Multi-GPU inference Run inference workloads that need several GPUs, such as the multi-GPU inference described in Verda’s 1XWM generative video collaboration. ### Generative video workloads Supply GPU infrastructure for video-generation systems where higher video quality is part of the application workflow. ### Distributed AI jobs Use an instant multi-node cluster when an AI workload must run across multiple machines rather than a single GPU instance. ### Managed research or engineering environments Provision a cluster with Slurm or Kubernetes when a team needs an established scheduler or container-orchestration environment without setting up the cluster manually. ## Perguntas frequentes ### How quickly can a GPU instance be provisioned? The GPU instances page states that on-demand instances can be provisioned in under a minute. ### How quickly can a multi-node cluster be provisioned? Instant clusters are described as provisioning in 20 minutes. ### Which cluster management options are available? Verda offers managed Slurm or Kubernetes for its instant clusters. ### What GPU models are listed? The source lists NVIDIA B300, B200, H200, H100, and A100 options, among others. ### Does Verda publish GPU pricing? The source identifies pay-as-you-go pricing for GPU instances, but the available material does not include specific rates or plan limits. ## Informações rápidas - **Category**: AI cloud infrastructure - **Primary resources**: NVIDIA GPU instances and multi-node GPU clusters - **Instance range**: 1x–8x GPU configurations - **Cluster software**: Managed Slurm or Kubernetes - **Provisioning**: GPU instances in under a minute; instant clusters in 20 minutes - **Pricing model**: Pay as you go for GPU instances