MC EXCLUSIVE Intel in talks to join DLI 2.0, backs India AI Mission with push to democratise AI

FILE PHOTO: Illustration shows words "Artificial Intelligence AI

• Intel plans to join India's DLI scheme under ISM 2.0 soon
• DLI 2.0 aims to boost chip design and create 50 fabless firms
• Intel supports India's AI mission and local engineering talent

Did our AI summary help?

Intel is in talks with the government to participate in the upcoming design linked incentive (DLI) scheme under  India Semiconductor Mission (ISM) 2.0, India president Gokul V Subramaniam has said, adding the US-based chipmaker will step in once the programme is opened to global players.

“We have been talking [with the government], and as soon as they open things up, we will be playing there,” Subramaniam, who is also the vice president of Client Computing Group at Intel Corporation, told Moneycontrol.

Participation in DLI would strengthen not just chip design but the broader electronics ecosystem in India, he said.

“If we can increase demand and consumption of electronics in India, naturally the supply chain and silicon design ecosystem benefits. It’s not just chip design — it’s the entire ODM ecosystem,” he said, referring to manufacturers building laptops, servers, data centres and network equipment for OEMs and hyperscalers.

The government is finalising the second phase of the DLI scheme to focus on six areas — compute, radio frequency, networking, power management, sensors and memory — to boost indigenous semiconductor design and enable India to manufacture up to 80 percent of electronic systems across sectors.

DLI 2.0, part of the next phase of ISM, also aims to support the creation of 50 fabless semiconductor companies and significantly expand India’s share in global chip design. Electronics and IT Minister Ashwini Vaishnaw has said the broader ISM 2.0 programme will include a push to advance India’s fabrication capabilities to 2nm by 2032.

Launched in December 2021, DLI 1.0 has already approved 24 startups, of which 18 have taped out their chip designs.

The government credits the progress to support in the form of EDA tools, IP access and multi-project wafer services. Over the past four years, about 67,000 students have also been trained in semiconductor technologies as part of a long-term talent development plan.

Push to democratise AI

Alongside its chip ambitions, Intel is positioning itself as a key contributor to India’s AI Mission, with a focus on making AI compute affordable and widely accessible.

“We’ve been focusing on heterogeneous, affordable AI compute. It’s not a one-size-fits-all model. Whether it’s cloud, network, edge, or the end-user PC — our approach varies,” Subramaniam said.

On PCs, Intel’s XPU architecture integrates CPU, GPU and NPU on a single chip, enabling varying levels of performance for training, inferencing, large language models and multimodal models.

“The idea is to take models and run them on the right compute needed, rather than build compute first and then try to fit workloads into it. That’s the whole thought process — heterogeneous and affordable AI,” he said.

Intel is working closely with startups, independent software vendors (ISVs), universities, academia and government bodies.

“We provide compute and collaborate closely. Even in Indic language translation and transcription, we’ve been working across the board — including with government-led initiatives. They use our compute and run their models, enabling them to operate across more than one compute platform,” he said.

Intel has had an engineering presence in India for 27 years, which continues to play a significant role in its global roadmap.

“We pretty much have teams working on every aspect of the compute platform — be it data centre, cloud, network, edge, or the PC client side,” Subramaniam said. The work spans processor design, IP, firmware, software, platform hardware, boards, reference designs and post-silicon validation.

“We’ve always stayed invested and continue to work on all critical programs on Intel’s roadmap. We also have teams working on custom silicon and some foundry areas as well,” he said.

On advanced node development, he said Intel’s Core Ultra Series 3 chips are based on 18A (1.8 nanometer) technology, with Indian teams contributing to the programme over the last couple of years.

When asked if Intel plans to expand R&D headcount in India, Subramaniam said it would depend on business conditions. “The key thing we look for is capability and depth of engineering talent. Intel has always believed in India for its engineering talent, as well as for the market. How much we grow or reduce is really a function of the business environment,” he said.

Skilling and ecosystem partnerships

Under India’s semiconductor push, Intel is also working with ISM on skill development across chip design, advanced packaging, testing and manufacturing.

The company has set up AI PC experience centre labs in collaboration with academic institutions, providing access to Xeon processors and client CPU-GPU platforms.

“It’s a very holistic approach. Building semiconductors isn’t just electrical engineering — it involves thermal, mechanical, materials science and packaging. All of these aspects are being looked at. It’s truly an exciting time for India,” Subramaniam said.

Intel works with ISV startups building AI solutions across productivity, education, healthcare and BFSI, providing platform support and performance optimisation, especially as applications are ported to x86 architecture. Funding decisions are handled globally by Intel Capital, though the India team provides inputs.

On enterprise AI deployment, Subramaniam said compute choice depends on workload. Some enterprises can run AI models on existing Xeon CPUs, especially for non-real-time analytics. Others may require higher performance locally, leveraging integrated CPU, GPU and NPU architectures.

This year, he expects AI prototypes to translate into meaningful, scaled applications across banking, agriculture, education and healthcare. “In India, the scale of data, diversity of datasets and the opportunity in Indic language translation make this a particularly exciting space,” he said.