QubitPage Technical Dossier

QubitPage Foundation

QubitPage Foundation — the GPU cloud platform powering SENTINEL AI training, fleet management, and investor datacenter credits. BYOK AI, VS Code integration, CLI/API, and NVIDIA DGX A100 infrastructure.

PlatformQubitPage®
GPU FleetDGX A100
Credit Ratio1:1 EUR
AccessInvestor Priority

1 — What is QubitPage Foundation?

QubitPage Foundation is the cloud control plane and GPU compute infrastructure that powers SENTINEL's AI training pipeline, fleet OTA updates, and investor-facing datacenter services. It is both the operational backbone of SENTINEL manufacturing and a standalone revenue-generating platform.

2 — Architecture

  • BYOK AI: Bring your own model — deploy, train, and serve custom AI workloads on NVIDIA A100/H100 hardware.
  • Resource Catalogue: On-demand GPU hours, storage, and networking — billed in credits.
  • VS Code Integration: Native extension for remote GPU sessions, Jupyter notebooks, and model monitoring.
  • CLI / API: Programmatic access for CI/CD pipelines, batch training jobs, and fleet inference deployment.
  • OTA Staging: All SENTINEL firmware and SENTINEL-OS updates are staged, validated, and deployed through QubitPage Foundation infrastructure.

3 — Investor Credits

Every euro invested in SENTINEL earns 1:1 datacenter credits on QubitPage Foundation. Credits can be used for:

  • GPU compute hours for AI model training and inference
  • Storage and data pipeline services
  • Priority queue access to NVIDIA DGX A100 / H100 nodes
  • Transfer within the QubitPage Foundation ecosystem

Credits never expire and are tracked in your Investor Portal.

4 — GPU Datacenter Specs

ComponentSpecification
Compute NodesNVIDIA DGX A100 — 8x A100 80 GB per node
InterconnectNVLink 600 GB/s + InfiniBand HDR 200 Gb/s
StorageNetApp AFF A800 — NVMe all-flash, 100 TB usable
NetworkDual 100 GbE uplink, BGP peering
Power200 kWp solar + grid backup (net-zero target)
LocationOn-site at SENTINEL factory, Ciurila, Cluj, Romania

5 — Use Cases

  • SENTINEL R&D: Training perception models, reinforcement-learning locomotion policies, and natural language interfaces for SENTINEL-OS.
  • Investor AI workloads: Investors with credits can deploy their own ML training jobs — LLM fine-tuning, computer vision, scientific computing.
  • Fleet Management: Real-time telemetry ingestion from deployed SENTINEL units, predictive maintenance models, and OTA rollout orchestration.