Enterprise-grade KVM virtual machines with dedicated NVMe SSD, isolated RAM, and full root access. Deploy in Mumbai, Bangalore, or Hyderabad โ ready in under 60 seconds.
8 Linux Distributions โ Pre-installed in Seconds
Every CruvzHub Linux VPC ships with enterprise features โ no extras to unlock.
All plans include root access, DDoS protection, snapshots, IPv4+IPv6, and 24/7 support.
Prices exclusive of 18% GST ยท No setup fees ยท Free migration ยท Cancel anytime
No contracts. Full root access. Cancel anytime. Plans from Rs.2,263/mo.
Windows Server 2019 and 2022 with full RDP access, Microsoft license included, and dedicated NVMe SSD. Run IIS, SQL Server, Active Directory, and .NET workloads with confidence.
Microsoft Windows Server license, RDP, DDoS protection, snapshots, and 24/7 support โ all included.
Prices exclusive of 18% GST ยท Microsoft Windows Server license included ยท No setup fees
Full RDP ยท Microsoft license included ยท NVMe SSD ยท 99.9% SLA
H100, A100, A40 โ purpose-built for training LLMs, running inference at scale, scientific simulation, and GPU rendering. CUDA 12.x, NVLink, NVMe-backed storage.
From LLM training to game server rendering โ CruvzHub GPU instances handle any compute-intensive task.
Enterprise-grade NVIDIA datacenter GPUs. Every instance is dedicated โ no shared GPU memory.
All GPU instances include dedicated GPU hardware, NVMe SSD storage, dedicated CPU + RAM, 1 Gbps uplink, and 24/7 support.
| GPU Hardware | NVIDIA H100 SXM, A100 PCIe (80GB), A40 (48GB) |
| GPU Memory | HBM3 (H100) or HBM2e (A100) โ up to 80 GB per GPU, 320 GB in 4x config |
| CUDA Cores | Up to 16,896 (H100 SXM). Full CUDA 12.x support |
| CPU Architecture | AMD EPYC / Intel Xeon โ dedicated cores per GPU instance |
| Storage Type | NVMe SSD โ fast data loading for training pipelines |
| NVLink (H100) | NVLink 4.0 โ 900 GB/s GPU-to-GPU bandwidth on multi-GPU configs |
| CUDA Libraries | cuDNN, cuBLAS, NCCL, cuSPARSE, TensorRT, RAPIDS |
| Container Support | NVIDIA NGC base images โ PyTorch, TensorFlow, JAX, Triton ready |
| Network | 1 Gbps uplink, HDR InfiniBand on cluster configurations |
| Operating System | Ubuntu 20.04/22.04 CUDA edition, or custom OS |
| Monitoring | GPU utilisation, memory, temperature โ real-time in portal |
| SLA | 99.9% monthly uptime โ contractual |
H100 ยท A100 ยท A40 ยท Dedicated GPU memory ยท CUDA 12.x ยท NVMe storage
Contact our GPU sales team for pricing, cluster configurations, and spot instance availability.