According to www.livemint.com, Indian semiconductor and physical-AI startups raised $523.3 million across 44 deals in 2025 — a sharp increase from $114.4 million across 48 deals in 2024 — yet remain critically underfunded for the transition from chip design to commercial manufacturing.
Funding Funnel Narrows Sharply After Seed Stage
Tracxn data reviewed by Mint shows a pronounced funding bottleneck beyond early-stage capital. From 2023 through August 2026, the sector recorded 94 seed and angel deals but only 20 Series A deals. Seed and angel funding totaled $44.1 million across 32 deals in 2024, dropped to $33.4 million across 28 deals in 2025, and further declined to $19.2 million across 15 deals through 10 August 2026. Series A activity, while improving, remains thin: four deals worth $25 million in 2024, six deals worth $53.6 million in 2025, and eight deals worth $72.9 million in 2026 to date.
Design Success ≠ Commercial Readiness
Government support has accelerated design-phase progress. As of June 2026, at least 24 projects were supported under India’s Design Linked Incentive scheme, and 105 companies gained access to advanced chip-design tools. The government confirmed that 23 chip designs had completed tape-out — the stage where final designs are sent to foundries for silicon fabrication. However, tape-out is only the beginning of the commercialization journey: chips must return from foundries, undergo packaging and testing, integrate reliably with software and hardware, and be qualified by customers before sale.
The $6.3 Million First-Silicon Gap
A Bengaluru-based founder of an AI processor startup — speaking anonymously — described how capital exhaustion occurs just before physical validation. The startup had completed simulation and programmable prototype testing but required $6.3 million for an initial batch of ~100 chips to verify post-manufacturing functionality, yield, and software-hardware compatibility. Unable to secure that sum, the company split its original ask:
“We initially sought $40–50 million to take the product through tape-out, early production and its next version, but now we have split that requirement into smaller rounds, starting with $15 million. That should fund tape-out and an initial run of about 1,000 chips.” — Anonymous founder, AI processor startup
Investor Hesitation Rooted in Capital Arithmetic
An investment banker advising deep-tech firms noted investors demand credible development plans, identified customers, and clear routes to market before committing large sums — especially pre-tape-out. Hemant Mohapatra, partner at Lightspeed and former AMD executive, emphasized the cost asymmetry:
“A first tape-out at a leading node can cost $30–40 million, while a tape-out on a mature 160nm process may cost $150,000–200,000.” — Hemant Mohapatra, partner at Lightspeed
He added that frontier chip ventures may require $500–750 million — far exceeding the capacity of conventional Indian venture funds. This creates a bimodal landscape: capital flows to either globally competitive AI-chip developers or low-cost chipmakers for appliances and EVs — leaving mid-tier startups stranded.
Post-Foundry Hurdles Multiply Complexity
Financing challenges persist after silicon returns from the foundry. Packaging, testing, and integration are typically handled by outsourced semiconductor assembly and test (OSAT) providers — services many Indian startups lack reliable access to. Without OSAT support, even functional chips cannot become shippable products. Scaling requires yield validation, component qualification, and diversified supplier networks. Shashwath T.R., co-founder and chief executive of Mindgrove Technologies, stressed that securing foundry and OSAT capacity must be planned months in advance — not treated as an afterthought.
Source: livemint.com
Compiled from international media by the SCI.AI editorial team.
