FlexAI's Commitment to India's Semiconductor Ecosystem

The Indian government's announcement of a new $15 billion incentive package to boost semiconductor manufacturing has sparked interest among global and domestic tech players. In this exclusive interview with Dataquest, Brijesh Tripathi, CEO of FlexAI, sheds light on how this initiative could align with FlexAI's vision of a "Universal AI Compute" landscape. Known for making AI more accessible and less dependent on specific hardware architectures, 

This innovative approach could complement India’s ambitions to become a global semiconductor hub. Tripathi discusses potential collaborations with Indian semiconductor manufacturers and how advancements in local semiconductor technologies could enhance FlexAI's ability to deliver optimized AI compute solutions, driving India's AI capabilities and supporting sustainable growth in the tech sector.

FlexAI's AI Compute Solution Aligns with India's Semiconductor Ambitions

Focus on how FlexAI's innovative approach to AI compute infrastructure could align with India's aspirations to become a global semiconductor hub.

Our approach to artificial intelligence offers significant flexibility by not restricting users to specific hardware architectures, thereby simplifying the model training process through its advanced software product. This flexibility aligns with India’s objective to make AI more affordable and accessible, fostering the development of innovative solutions across various sectors.

As the demand for AI compute power is projected to grow exponentially over the next decade, India must explore options to increase AI compute capacity. One effective strategy is to develop indigenous solutions, including new hardware, chips, and systems, to reduce dependency on single sources. FlexAI’s AI Cloud service further supports this initiative by enabling new entrants to participate in the AI compute supply chain without the complexities typically associated with integrating new hardware.

By providing accessible and cost-effective AI tools, FlexAI helps drive India’s ambition to expand its AI capabilities and support sustainable growth in the technology sector.

How does FlexAI's vision for "Universal AI compute" fit into India's recent $15 billion semiconductor push? Do you see opportunities for collaboration with Indian semiconductor manufacturers or tech giants?

India's semiconductor ecosystem is evolving rapidly, with multiple OSAT (Outsourced Semiconductor Assembly and Test) units being established, while EMS (Electronics Manufacturing Services) companies, which have had a longer presence in the country, are advancing up the value chain by setting up OSAT plants.

This creates a strong opportunity for collaboration with the local ecosystem to develop innovative hardware products. However, building and growing such an ecosystem requires time and strong collaboration.

As India continues to develop its semiconductor capabilities, including silicon manufacturing and OSAT infrastructure, We can seize this opportunity to partner with key OSAT and silicon manufacturing players to develop its next generation of systems and chips. These partnerships will be crucial for advancing hardware innovation in this rapidly expanding ecosystem.

As India aims to become a global player in semiconductor manufacturing, how do you foresee this impacting the AI compute landscape, especially in terms of availability, cost, and efficiency of high-performance chips for AI workloads?

Currently, Indian semiconductor manufacturing is focused on mature nodes, which are not suitable for AI chips that require advanced nodes—an area not yet planned for. However, the landscape is expected to change when it comes to semiconductor packaging and system integration.

As India works to increase AI compute capacity by developing indigenous solutions, including new hardware, chips, and systems, the risks associated with supply chain dependencies will decrease. In addition to gaining greater control over the supply chain, India will also be in a stronger position to influence the cost of building this new infrastructure, further enhancing its competitiveness in the AI and semiconductor sectors.

FlexAI focuses on re-architecting compute infrastructure at a system level. Could you elaborate on how advancements in semiconductor technology could enhance FlexAI's ability to deliver optimized AI compute solutions?

New packaging technologies open up the possibility of exploring different architectural options, while advancements in thermal and power delivery technologies improve efficiencies and enhance system-level capabilities.

Innovations in chiplets, packaging, system integration, power delivery, and cooling options will continue to push the boundaries of what we can offer, both through its datacenter-in-a-box solution and FlexAI Cloud Services (FCS) as a cloud service. By pushing semiconductor technologies to their limits, we are committed to building the best possible products.

Are you open to partnering with Indian semiconductor fabs or startups to develop custom AI chips optimized for FlexAI’s infrastructure? What would be the key considerations for such partnerships?

We currently plan to architect our chips internally while partnering for their design, with a focus on selecting the best IP for our chips. Key considerations include companies that offer scalable and power-efficient IP. Additionally, we are open to partnering with startups and fabs to develop AI chips and solutions for our FCS product.

This aligns perfectly with our strategy to enable more AI hardware options, as FCS delivers heterogeneous AI compute without added complexity, making it an ideal platform for innovative AI solutions.

Given the challenges India faces in establishing a robust semiconductor ecosystem, such as supply chain logistics and skilled workforce development, how can companies like FlexAI contribute to addressing these issues?

Many of FlexAI’s employees come from diverse technical backgrounds, with experience in multinational semiconductor companies. In India, institutions like IITs and IISc have initiated programs focused on similar fields. FlexAI encourages its employees to collaborate with professionals from these institutions to conduct training, seminars, and guide graduate students in research projects.

As part of its commitment to talent development, FlexAI has already onboarded interns in India and actively supports internship programs.

As India works towards developing indigenous AI compute solutions, including new hardware, chips, and systems, the risk of supply chain dependencies will reduce. With greater control over the supply chain, India will also have more influence over the cost of building this new infrastructure, positioning itself as a key player in the development of AI technologies.

How might India’s evolving policies in semiconductor manufacturing and AI development influence FlexAI’s strategy for expanding its operations or R&D activities in India?

India is taking the right steps to develop its semiconductor ecosystem, driven by the growth of AI and the broader digital space. FlexAI sees this as a great opportunity and is already building a team in India to support its operations.

In collaboration with MeitY, FlexAI is actively seeking partnerships with top academic institutions and technology companies to further its goals.

Our mission of delivering more compute with less complexity requires global talent and expertise, and India plays a key role in this strategy. The company is expanding its presence in India to develop AI-specific chips and systems, leveraging favorable public policies and government incentives aimed at fostering deep tech startups. These initiatives align with FlexAI’s vision to grow its capabilities while contributing to India’s emerging semiconductor landscape.

With the new incentives focusing on advanced technology investments, how do you see the future of AI compute evolving in terms of hardware optimization, especially with high-value semiconductor fabs potentially setting up in India?

The AI space is evolving rapidly, but the setup and production of semiconductor fabs is a multi-year journey. While inviting advanced technology investments is a positive step, the near-term impact on hardware optimization will likely be minimal.

In this context, it is crucial for nation-states to develop their technology and manufacturing capabilities, as supply chain control and technology export policies will increasingly influence political and economic stability. Countries like India must focus on building a strong technology and manufacturing roadmap to become more self-reliant. Over the next few years, we expect the development of new hardware solutions that can be manufactured globally and, eventually, within India, contributing to long-term technological resilience.

How do you think India's entry into the advanced semiconductor market could reshape the global landscape for AI compute? Would this influence your approach to developing AI compute solutions that leverage Indian semiconductors?

We are closely watching the developments in India's technology landscape. With its large consumer market and strategic geopolitical position, India is well positioned to make significant contributions, provided it can execute plans in a timely manner. While the short-term impact may be limited due to the long lead times required to build the necessary ecosystem, the long-term opportunities are promising.

FlexAI’s aim of delivering more compute with less complexity requires global talent and expertise, and we are partnering with startups and OSATs in India to develop AI chips and solutions for our FCS product. Our goal is to build an ecosystem that fosters the development of more AI hardware options, leveraging FCS to provide heterogeneous AI compute without added complexity. This approach positions us to take advantage of India’s growing role in the AI and semiconductor sectors over the long term.

From your perspective, how critical are advancements in semiconductor technologies for creating a more efficient and universally accessible AI compute infrastructure?

Over the years, the cost of enabling new technologies has increased on average. While technological advancements are essential, the focus must also include innovative architectural ideas and co-development with software optimization to provide accessible solutions.

Our goal is to deliver significantly improved Total Cost of Ownership (TCO) to FCS customers in the coming years by leveraging technologies that enhance both the cost efficiency and power of AI solutions. We will continue pushing semiconductor technologies to their limits to build the best version of our Datacenter-in-a-box solution, combining cutting-edge advancements with innovative architectural approaches.

What is your vision for the next 5-10 years in the context of semiconductor development and AI compute infrastructure, and how might India’s growing role in the semiconductor industry influence that vision?

FlexAI aims to make AI compute less complex, more accessible, and economical. By leveraging a larger ecosystem of capable partners and access to advanced technology and resources, we are confident in our ability to realize this vision.

Our goal is to provide more compute while reducing the complexity of integrating new architectures and solutions into the AI compute supply. Over the next few years, we plan to scale the FCS product to support customers who face challenges in accessing AI compute due to supply or cost limitations. Looking ahead to the next 5-10 years, we aim to develop a full-stack solution, including AI-optimized chips and systems, to improve the total cost of ownership (TCO) for our FCS customers.