A new semiconductor startup founded by engineers Bragadeesh and Lannan Jiang has launched with the goal of lowering barriers to custom chip development, addressing cost and timeline challenges that have kept application-specific silicon out of reach for most companies.

The venture enters a market where custom chip design typically requires millions in upfront investment and development cycles measured in years. By contrast, the founders claim their platform will compress iteration timelines and reduce financial thresholds, though specific pricing and performance metrics have not been disclosed.

Why Custom Silicon Matters

Application-specific integrated circuits (ASICs) - chips designed for particular workloads rather than general computing - can deliver substantial performance and efficiency gains over off-the-shelf processors. Major technology companies including Google, Amazon and Tesla have invested heavily in custom silicon for data centres and specialised applications. Smaller firms have largely been priced out of this approach, relying instead on configurable chips that offer less optimisation.

The economics have been prohibitive. Traditional ASIC development involves semiconductor design firms, fabrication partnerships and extensive validation cycles. A single chip design can cost between $30 million and $80 million according to industry analysts, with no guarantee of commercial success.

Technical Approach

Details of the startup's technical methodology remain limited in public statements. The founders indicated their platform focuses on design automation and tooling that simplifies the creation of custom architectures. This suggests an approach similar to other recent efforts to abstract away low-level hardware description languages and physical design constraints.

The semiconductor industry has seen related initiatives. Companies like Efabless have offered open-source design flows, while SiFive commercialised customisable processor cores. Academic projects including the OpenROAD toolchain have tackled design automation. The new startup's differentiation and technical advantages over these existing efforts have not been specified.

Market Timing

The launch comes during a period of renewed interest in domain-specific architectures. As Moore's Law improvements slow, companies across sectors have explored custom silicon to maintain performance growth. The artificial intelligence boom has particularly accelerated this trend, with machine learning workloads demonstrating clear benefits from specialised hardware.

Venture capital has followed. Semiconductor design automation and custom chip startups raised over $2 billion in 2023 according to PitchBook data, though that figure includes established players and represents a decline from 2021 peaks.

Challenges Ahead

The startup faces substantial technical and commercial obstacles. Chip design automation has proven difficult to commoditise despite decades of effort. Each new process node introduces complexity that can undermine abstraction layers. Fabrication partnerships remain expensive and capacity-constrained, particularly at advanced nodes.

Customer acquisition presents another hurdle. Companies considering custom silicon must weigh substantial switching costs and integration challenges against potential performance gains. Building a customer pipeline requires both technical credibility and business development resources.

The founders have not disclosed funding details, team size or customer commitments. Without additional information about their technical approach, go-to-market strategy or financial backing, assessing the venture's viability remains speculative.

The company plans to announce further details about its platform and initial partnerships in coming months.