ESDS Software Solution Limited has introduced a new sovereign-grade GPU-as-a-Service offering, unveiled during the company’s 20th Annual Day. This service aims to address the increasing computational demands of AI, ML, GenAI, and LLM workloads across a diverse spectrum of sectors, including enterprises, BFSI, research institutions, and government agencies.
The launch positions ESDS as a comprehensive provider, extending its portfolio to include large-scale, sovereign-grade GPU infrastructure in addition to its existing cloud, managed services, data centre infrastructure, and software solutions. The company emphasized that this new offering is designed to deliver high-performance AI compute capabilities at a global scale.
This announcement coincides with projections that global spending on AI-optimised servers, which encompasses GPUs and accelerators, is set to reach $329.5 billion by 2026. The growth is largely attributed to the rising demand for deterministic, high-throughput computing environments.
According to ESDS, the new platform will enable organizations to efficiently run mission-critical AI workloads on purpose-built GPU SuperPODs, which are engineered for secure operations, consistent performance, and low-latency distributed training. The company has refined its expertise into a fully managed GPU infrastructure stack, aimed at enabling organizations to scale their AI initiatives on a reliable architectural foundation.
Piyush Somani, the promoter, managing director, and chairman of ESDS, stated that this launch responds to the surging demand for extensive AI infrastructure. “With this launch, we are democratising access to large-scale GPU clusters and SuperPODs, making them straightforward, transparent, and purpose-built for enterprises that have AI ambitions,” Somani noted. He further asserted that the ESDS GPU SuperPODs “fundamentally change that narrative by delivering predictable performance, stability, and scale.”
To further empower businesses, ESDS has developed the SuperPOD Configurator tool, allowing enterprises to select their GPU model, design their cluster, and gain immediate visibility into architectural and cost parameters. The core of the offering includes a range of high-performance GPU systems, featuring NVIDIA DGX and HGX models such as B200, B300, GB200, NVL72, along with AMD’s MI300X platforms. These systems are tailored to support extremely large model training, enhance inference workloads, execute simulations, and manage substantial clustered data operations.
ESDS further reported that its GPU SuperPODs utilize high-bandwidth NVLink, unified memory pools, intelligent scheduling, enhanced thermal management, and AI-tuned orchestration to ensure consistent performance at any scale. The service portfolio encompasses consultancy for captive GPU clusters, deployment of GPU environments, dedicated GPU infrastructure-as-a-service, hybrid CPU+GPU cloud options, and a fully managed on-demand GPU cloud. ESDS will oversee architecture design, network optimisation, container orchestration, performance tuning, and round-the-clock monitoring with AI/ML Ops support.
A significant aspect of the rollout is the SuperPOD Configurator, which aids enterprises in designing their AI infrastructure by selecting GPU models, compute density, memory profiles, storage tiers, and interconnect options. The system automatically generates optimized architectures, performance estimates, and cost projections.
ESDS referenced a case study involving a research lab that successfully reduced the training time for a 50-billion-parameter model from over 40 days to just 10 days, cutting costs by 60% and achieving 30 times faster inference after transitioning to NVL72-based GPU systems paired with optimized containers and high-speed NVLink.
The company concluded that its offering is designed to meet global AI performance standards while being optimized in India. Currently, ESDS services over 1,300 clients across enterprises, BFSI, and government sectors, providing transparent pricing, flexible consumption models, and integrated cloud and managed services.
See also
Global Processor Market Set to Reach $210.26B by 2033, Driven by AI and Cloud Demand
One Stop Systems Launches PCIe 6.0 Products, Valuation Estimate at $8 Amid 156% Shareholder Return
Google’s Gemini 3 Launch Transforms AI Landscape, Challenging Nvidia’s Dominance
Google CEO Pichai: Quantum Computing Approaching Major Breakthroughs in 5 Years
Q.ANT Launches NPU 2.0, Promising 30x Lower Energy Use and 50x Higher Performance
