AI & HPC infrastructure.

Explore AI workstations, GPU servers, HPC clusters, storage and networking with Curobotic in Guwahati. Each system is configured around your applications, users and facilities, with hardware, deployment and support clearly itemised in a tailored quotation.

NVIDIA GB200 compute board with processors, liquid-cooling pipes and interconnects on display
NVIDIA GB200 compute platform Photo credit

Explore system categories

From a workstation. To a shared cluster.

10 product familiesConfigured around your workload.
Pricing by quotation.

NVIDIA RTX 6000 Ada workstation graphics cards arranged on a studio workbench
NVIDIA RTX 6000 Ada GPUs Photo credit

AI & GPU workstations

A dedicated workstation gives a researcher, developer or designer direct access to local computing resources. Curobotic scopes AI and GPU workstations around the applications you use, the space available and the support you need after installation.

Four NVIDIA H100 GPU accelerators installed side by side in a development system
NVIDIA H100 multi-GPU system Photo credit

AI development servers

Give a development team a shared place to prepare data, run experiments and integrate AI applications. Curobotic scopes AI development servers for the users and project environments that need access, with a clear configuration and deployment brief.

NVIDIA GB200 board showing two Blackwell GPUs, a Grace CPU and cooling connections
NVIDIA Grace Blackwell compute Photo credit

Multi-GPU training servers

Bring several accelerators into a server configuration designed for a defined training workload. Curobotic scopes multi-GPU training servers with attention to device compatibility, host resources, interconnects and the facilities needed to operate the system.

Cisco UCS chassis populated with rack-mounted blade server nodes
Cisco UCS compute chassis Photo credit

CPU compute nodes

Add CPU computing capacity for simulations, analysis and other workloads that need it. Curobotic scopes compute nodes around application behaviour, memory requirements and integration with an existing or planned research computing environment.

Summit supercomputer cabinets showing rack doors, cabling and the adjacent compute row
Summit HPC cluster racks Photo credit

Configured HPC cluster packages

Procure a defined computing environment with its supporting components accounted for. Curobotic’s configured HPC cluster packages bring compute, shared services, networking and delivery scope into one reviewed bill of materials for the intended research workload.

NVIDIA H100 PCIe GPU accelerator photographed on a white workbench against a dark background
NVIDIA H100 PCIe GPU Photo credit

AI inference servers

Run trained models as part of a controlled application service. Curobotic scopes AI inference servers around the model, request pattern and operating requirements, with hardware and serving setup defined against the experience your application needs to deliver.

Front panels of Dell PowerEdge servers mounted together in a rack
Dell PowerEdge rack servers Photo credit

Enterprise rack servers

Choose rack-mounted computing resources for the applications your organisation operates. Curobotic scopes enterprise rack servers for virtualisation, databases, institutional systems and other workloads, with configuration and support terms made explicit in the quotation.

Metal network-attached storage enclosure with a vertical array of removable drive bays
Multi-bay network storage Photo credit

NAS, SAN & parallel storage systems

Store and deliver data in the form your applications need. Curobotic scopes NAS, SAN and parallel storage systems for business and research environments, comparing access methods, usable capacity and operating requirements before a configuration is proposed.

Fibre patch panels with rows of blue connectors inside a network rack
Data-centre fibre connections Photo credit

High-speed networking equipment

Connect computing resources with a network selected for the traffic they must carry. Curobotic scopes high-speed networking equipment for server, storage and research environments, accounting for endpoints, cable paths and management as part of the configuration.

NVIDIA Jetson AGX Orin developer kit beside its NVIDIA packaging
NVIDIA Jetson AGX Orin Photo credit

Edge AI devices

Place suitable inference capacity near a camera, machine or local application. Curobotic scopes edge AI devices according to the model workload, input interfaces and installation environment, with software and deployment work distinguished from equipment supply.

From equipment to a working environment

Plan the complete project.

How are AI and HPC systems priced?

Pricing is quote-based. We first review the workload, configuration, facilities and delivery scope. The proposal identifies manufacturer and model details, components, software licences, installation, warranty and support. Availability, lead time and commercial terms are confirmed in the quotation rather than inferred from a generic starting price.

Can equipment supply and installation be quoted separately?

Yes. A quotation can distinguish hardware supply from site preparation, software deployment, migration, training and ongoing administration. This helps your technical and procurement teams compare equivalent scope and decide which work will be handled by your organisation.

Where can Curobotic scope infrastructure projects?

Our team is based in Guwahati, Assam. We scope projects across Assam, Itanagar and other Arunachal Pradesh locations, and Shillong, with remote preparation and on-site work agreed for each engagement. Delivery access, facilities readiness, travel and support coverage are confirmed before a project is accepted.

Your next chapter starts here.

Your workload. The right configuration.

Tell us about your applications, users and delivery location. We will scope the equipment, installation and support required for your project.

admin@curobotic.com