High Performance Computing (HPC) the Future of Manufacturing in India

What are the challenges in accelerating High-Performance Computing (HPC) in India? How do you resolve these challenges? 

Accelerating High-Performance Computing (HPC) in India faces several significant challenges. One of the primary obstacles is the limitation in infrastructure. HPC requires high-density compute systems that balance performance, energy efficiency and serviceability. However, many institutions and industries lack access to the latest infrastructure, often relying on outdated or insufficient systems. Even when advanced solutions are available, the high operational costs deter widespread adoption. For example, systems employing liquid cooling, though highly energy-efficient, can still be cost-intensive to implement and maintain​.

Another major challenge is the skill gap in managing and optimising HPC environments. Advanced setups demand expertise in cluster management, parallel programming and optimisation, which are not yet mainstream in Indian academic curricula. As a result, the lack of skilled professionals becomes a bottleneck in fully leveraging HPC capabilities. To bridge the skills gap, HPC education must be embedded in academic curricula, supported by training programmes in collaboration with industry leaders like C-DAC, Netweb and IITs, with incentives like scholarships and job placements to attract talent. Increasing R&D funding through higher government budgets, tax incentives and international partnerships will further accelerate growth, enabling India to compete with global HPC leaders

Thirdly, software optimisation for modern HPC architecture remains a challenge, as most existing software solutions are not designed for extreme parallelism and resilience, resulting in suboptimal performance.

Lastly, energy consumption is another critical issue. HPC systems consume significant amounts of power and scaling them up can put immense pressure on India's power infrastructure. Therefore, there is a pressing need to adopt energy-efficient systems and architectures to mitigate this challenge.

What are the prospects of manufacturing units coming up in India using HPC?

The prospects of manufacturing units leveraging High-Performance Computing (HPC) in India are highly promising, driven by a combination of technological advancements, government initiatives; and the country’s growing industrial and economic landscape. HPC offers transformative potential for manufacturing by enhancing design, simulation, optimisation and production processes.

Here’s an analysis of the key factors shaping these prospects –

Growing Demand for Advanced Manufacturing

HPC can enable manufacturers to –

Optimise Designs – Use simulations to test product designs virtually, reducing the need for physical prototypes and cutting costs.

Enhance Efficiency – Model production processes to minimise waste, improve energy efficiency and boost throughput.

Innovate Rapidly – Leverage HPC for advanced materials research, AI-driven automation, and predictive maintenance, giving Indian manufacturers a competitive edge globally.

Government Support and Infrastructure Development

The Indian government has recognised HPC as a critical enabler of industrial and scientific progress, exemplified by the National Supercomputing Mission (NSM) launched in 2015. Key aspects that support manufacturing prospects include –

Supercomputing Infrastructure – The NSM aims to deploy over 70 petaflops of computing power across the country, with indigenous server platforms already in use. This infrastructure can be accessed by manufacturing units for research and development.

Indigenous Technology – India has made significant strides in developing indigenous HPC technology, particularly for manufacturing applications. Netweb’s Make in India HPC systems are purpose-built to deliver high-performance computing capabilities tailored for manufacturing workloads such as simulation, modelling and digital twin analysis. Our Tyrone HPC servers and clusters leverage the latest CPU and GPU technologies, offering scalable configurations to meet diverse computational demands.

PLI Schemes – Production-Linked Incentive schemes in sectors like electronics and semiconductors encourage local manufacturing, where HPC can play a pivotal role in design and testing phases.

Do you invest in R&D in your organization? If yes, in what segments?

Netweb actively invests in Research & Development (R&D) across multiple technology segments, particularly in High-Performance Computing (HPC), AI, private Cloud and enterprise IT solutions, aligning with India’s digital transformation initiatives. The company focuses on developing high-density compute clusters, AI-optimised servers & workstations and private Cloud solutions like Skylus, enhancing performance for GPU-intensive workloads, deep learning and data analytics. Recently, we launched Skylus.ai, a GPU aggregation-disaggregation appliance designed to optimise GPU resource management for AI including GenAI workloads. It addresses the critical challenges faced by organisations in utilising multi-vendor GPU and CPU resources, offering a vendor-agnostic solution that drives faster ideation, fosters collaboration and accelerates experimentation while optimising resource utilisation and total cost of ownership.

Last year, we also inaugurated our flagship end-to-end high-end computing servers, storage and switch manufacturing facility. The facility signifies a comprehensive leap in manufacturing capabilities for cutting-edge computing systems. It encompasses the entire spectrum, from designing Printed Circuit Boards (PCBs) to surface mounting on PCBs and finally the production of complete systems, showcasing India's prowess in creating sophisticated technology products. It also demonstrates our steadfast commitment to fostering innovation and bolstering self-reliance in India's technology sector.

How reliable is AI-driven predictive maintenance?

AI-driven predictive maintenance (PdM) is a data-driven approach to anticipate machines and equipment failures and conduct pro-active repairs.  The factors that affect the reliability of AI-driven predictive maintenance are data quality, algorithm selection, human expertise, implementation and integration; and scalability. By successfully addressing these factors, organisations can derive the complete benefits of AI-driven predictive maintenance and reap significant improvements in their daily operations.

Certain challenges affect the reliability of PdM, including data inconsistency, sensor malfunctions and cybersecurity risks in Industrial IoT (IIoT) environments. AI models also require continuous retraining to adapt to new wear-and-tear patterns, ensuring sustained accuracy and performance. Despite these challenges, AI-powered predictive maintenance remains a game-changer for industries, significantly improving asset reliability, operational efficiency, and cost savings.

Real-world industrial settings require constant AI model improvements to maintain high accuracy. Although there can be a debate on the accuracy level in real-scenario, one thing is inevitable – AI-driven PdM is the future. With ongoing advancements in edge AI, federated learning, and cloud computing, AI-driven predictive maintenance will become even more precise, secure, and adaptive, making it a critical asset for modern industries.

Read more interviews here