Skip to main content

GPU Cloud Overview

NeuroHash GPU Cloud provides on-demand access to NVIDIA GPUs for AI training, inference, and research workloads. Launch instances in under 60 seconds, scale as needed, and pay only for what you use.

Key Features​

Instant Provisioning​

Launch GPU instances in seconds — no waiting for hardware allocation or manual provisioning. Select your GPU type, region, and SSH key, then you're live.

Pay-Per-Minute Billing​

Billing is calculated per-minute with hourly charges. No minimum spend, no upfront commitments, no egress fees. Payments are in ETH on Robinhood Chain, converted from USD pricing using live CoinGecko exchange rates.

Enterprise-Grade Hardware​

Access the latest NVIDIA GPU architectures:

  • B200 — Latest Blackwell architecture, 180GB HBM3e, 8 TB/s bandwidth
  • H200 — 141GB HBM3e memory, 4.8 TB/s bandwidth
  • H100 — 80GB HBM3, 3.35 TB/s bandwidth, the industry standard for LLM training
  • A100 — 80GB HBM2e, proven workhorse for training and inference
  • A10 — Cost-effective for inference and light training
  • GH200 — Grace Hopper superchip, 96GB HBM3
  • RTX A6000 — 48GB GDDR6, great for research and visualization

Multi-GPU Configurations​

Scale from a single GPU to 8-GPU instances:

ConfigUse Case
1x GPUDevelopment, fine-tuning, inference
2x GPUMedium training jobs, multi-model inference
4x GPULarge model training, distributed workloads
8x GPULLM training, massive parallel compute

Pre-Configured Environments​

Every instance is ready to use with:

  • CUDA toolkit (latest)
  • PyTorch & TensorFlow
  • Python 3.x with common ML packages
  • JupyterLab for interactive development
  • Full root access via SSH

High-Performance Networking​

  • Up to 3.2 Tb/s network throughput
  • 400 Gbps per GPU for multi-node training
  • NVIDIA Quantum-2 InfiniBand on cluster configurations
  • Non-blocking, full-bisection bandwidth

How It Works​

1. Fund wallet with ETH  -->  2. Select GPU type  -->  3. Launch instance
         |                           |                        |
    Robinhood Chain            Choose region           SSH in & train
    (Chain ID: 4663)          & config                  your models
  1. Fund your wallet with ETH on Robinhood Chain
  2. Select your GPU type, region, and SSH keys
  3. Launch — instance is live in under 60 seconds
  4. Connect via SSH and start your workload
  5. Terminate when done — billing stops immediately

Pricing​

GPU Cloud uses per-minute billing with no minimum spend. See Instance Types for detailed pricing or visit the Pricing page.

Next Steps​