GPU as a Service
On-demand GPU compute for AI model training, fine-tuning, inference and high-performance workloads — without the capex of owning the hardware.
- Training & inference
- Dedicated or shared
- Scale up on demand
From the first trench to the last switch port — we plan, design, procure, install, integrate, test, commission and then keep it all running.
From owned fibre to GPUs in the rack — we now deliver infrastructure as a service. Spin up compute, stand up an AI lab, digitise a courtroom or lease fibre, all from one accountable partner.
On-demand GPU compute for AI model training, fine-tuning, inference and high-performance workloads — without the capex of owning the hardware.
Turnkey AI & ML labs for universities, institutes and government — GPU workstations, servers, networking, software stack and hands-on enablement.
Courtroom digitisation — video-conferencing for hearings, digital displays, recording, e-filing and case-management infrastructure, networked and supported 24×7.
Provision virtual servers in minutes — Linux or Windows, right-sized CPU, memory and storage, with backups and monitoring built in.
~1000 km of owned fibre across West Bengal, Bihar & Jharkhand — dark fibre and bandwidth on lease, maintained for the full tenure. DoT IP-1 registered.
Enterprise storage, data management, backup and disaster recovery that keep critical data available.
Data-centre compute, virtualisation, IaaS, SaaS and cloud security — designed, deployed and managed.
On-demand cloud GPUs billed per GPU-hour. No long contracts — scale from a single L4 to multi-GPU B200 clusters.
| GPU | vRAM | vCPUs | RAM | On-demand | |
|---|---|---|---|---|---|
| NVIDIA B200Flagship | 192 GB | 32 | 400 GB | ₹664.05$6.99/hr | Request → |
| NVIDIA H200Popular | 141 GB | 30 | 375 GB | ₹379.05$3.99/hr | Request → |
| NVIDIA H100Popular | 80 GB | 26 | 250 GB | ₹255.55$2.69/hr | Request → |
| NVIDIA RTX PRO 6000 | 96 GB | 32 | 170 GB | ₹182$1.90/hr | Request → |
| NVIDIA A100 80GB | 80 GB | 16 | 115 GB | ₹189$2.10/hr | Request → |
| NVIDIA A100 40GB | 40 GB | 16 | 115 GB | ₹179$1.98/hr | Request → |
| NVIDIA A40 | 48 GB | 16 | 100 GB | ₹96$1.44/hr | Request → |
| NVIDIA L40S | 48 GB | 60 | 220 GB | ₹102$1.20/hr | Request → |
| NVIDIA A30 | 24 GB | 16 | 90 GB | ₹90$1.35/hr | Request → |
| NVIDIA L4Best value | 24 GB | 25 | 110 GB | ₹49$0.57/hr | Request → |
Prices are per GPU-hour for a 1× GPU instance, on-demand, exclusive of applicable taxes (GST). Monthly & annual plans, multi-GPU nodes, volume and enterprise pricing on request.
Complete turnkey solutions across a wide spectrum of technology infrastructure — detailed project planning, system design, equipment procurement, on-site installation, system integration, testing and final commissioning to ensure operational readiness.
LAN, WAN, wireless networking and mobile communication services for seamless communication and data transfer across locations.
APPLICATION — Interconnecting systems within offices and campuses, linking multiple locations, remote access, data & voice transmission.
Multimedia and audio-visual system solutions, video wall and display solutions.
APPLICATION — Conference rooms, auditoriums, classrooms and control rooms.
CCTV camera systems, fire alarm system services and automatic boom barrier systems.
APPLICATION — Monitoring premises, detecting fire incidents, generating alerts and regulating vehicle & personnel entry.
Customised outdoor rack solutions and equipment enclosures for harsh and outdoor environments.
APPLICATION — Protecting networking and communication equipment in exposed locations.
Compute (servers & processing), storage & data management, backup & recovery, IaaS, SaaS and cloud security.
APPLICATION — Storing, processing and managing enterprise data, hosting applications and securing cloud operations.
Servers, storage, virtualisation and the network fabric between them — the physical foundation that data-heavy and AI workloads depend on.
Maintenance and support for installed IT and communication infrastructure through AMCs and Facility Management Services — formal agreements for a specified tenure, extendable by mutual understanding.
Technical personnel on-site or remote — monitoring, administration and operational support of networks, servers and related infrastructure.
APPLICATION — Continuous operation.
Periodic inspection and routine servicing to catch issues early and keep equipment performing.
APPLICATION — Computers, printers, networking devices.
Fault identification and rectification, repair or replacement of defective components.
APPLICATION — Rapid issue resolution.
Continuous monitoring, technical assistance, breakdown and emergency response, vendor coordination.
APPLICATION — Minimise downtime.
Development, deployment and management of fibre infrastructure — a mix of execution revenue and long-term recurring income, riding the demand for high-speed digital connectivity.
Dark fibre or bandwidth capacity leased to telecom operators, enterprises and institutions for a defined period — with network maintenance throughout the tenure.
APPLICATION — ~1000 km own network in WB, Bihar & Jharkhand, leased to RailTel, PGCIL, SIFY, Tulip and their clients.
Civil works for underground fibre: trenching and digging for cable routes, duct installation, and restoration of roads and pavements.
APPLICATION — Laying and integrating underground fibre networks.
Registered as Infrastructure Provider Category-1 (IP-1) under the Department of Telecommunications, Government of India — to establish and maintain dark fibre, right of way, duct space and towers for lease, rent or sale to telecom licensees under Section 4 of the Indian Telegraph Act, 1885.