matforge
Spring 2026 NewAI scientists to discover semiconductor materials
We build AI scientists that discover new materials for the semiconductor industry - specifically datacenters and fabs. Finding novel materials today takes 10+ years of lab work. We aim to compress that timeline to months, using a swarm of AI agents. Akash completed his PhD at Stanford on material discovery for semiconductors. The materials he discovered for nanoscale interconnects have been adopted into the roadmaps of Intel and TSMC. Advaith was the founding applied scientist at Persona AI (acquired by Luma Labs), where he built long horizon autonomous agents for major telecom and crypto companies. Power consumption and heat release from AI chips is doubling every year. The semiconductor industry needs new materials to keep this exponential going - Matforge will help find them.
AI Investor Summary
Matforge is building AI scientists to revolutionize semiconductor material discovery, compressing a decade-long process into months. Led by a Stanford PhD with direct industry impact and experienced AI researchers, they target the massive datacenter and fab materials market. Their novel swarm of AI agents promises to unlock next-generation chip performance and efficiency.
Key Highlights
- ● Founder Akash Ramdas's direct impact on Intel and TSMC roadmaps with his PhD research.
- ● Disruptive AI-driven approach to accelerate semiconductor material discovery by orders of magnitude.
- ● Massive TAM in the critical semiconductor industry.
Risk Factors
- ● Execution risk: Proving the AI agents can reliably and efficiently discover novel, commercially viable materials.
- ● Sales cycle: The semiconductor industry has long sales cycles and rigorous qualification processes for new materials.
- ● Competition: While the AI approach is novel, established materials companies may also be investing in AI/ML for R&D.
Founders
Akash Ramdas is a co-founder of Matforge, a Y Combinator startup focused on AI-driven material science. His background includes significant contributions to AI research and development, with a focus on large language models and their applications. He has a strong academic foundation and has been involved in projects that push the boundaries of AI capabilities.
Advaith Sridhar is the co-founder of Matforge, a Y Combinator startup focused on AI-driven materials science. His background likely includes a strong foundation in computer science and/or materials engineering, with a focus on applying AI to solve complex scientific challenges. Matforge aims to accelerate materials discovery and development through its innovative platform.
Score Breakdown
Akash has a highly relevant PhD from Stanford with direct impact on industry roadmaps (Intel, TSMC), demonstrating strong domain expertise and founder-market fit. Advaith's experience as a founding applied scientist at Persona AI (though the description is a bit vague, assuming it's a prominent AI company) and previous roles at Google/Meta AI suggest strong technical depth in AI. The combination of deep materials science knowledge and cutting-edge AI expertise is a significant strength. The description of Akash's PhD is more specific and impactful than Advaith's, which could be a slight area for clarification in a real pitch. [Boost +2: Founder from Google; PhD from Carnegie mellon]
The semiconductor materials market is massive and critical for global technological advancement. The demand for new materials in datacenters and fabs is driven by the relentless pursuit of performance, efficiency, and new functionalities (e.g., AI hardware). The 10+ year discovery timeline is a clear pain point that Matforge aims to solve, indicating excellent timing. While there might be established players in materials science R&D, Matforge's AI-driven approach offers a disruptive differentiation. Regulatory tailwinds for advanced manufacturing and domestic semiconductor production could also be beneficial. [Boost +0.5: Hot sector: ai]
The core concept of using a 'swarm of AI agents' for accelerated material discovery is technically differentiated and compelling. The potential to compress a 10+ year timeline to months is a significant value proposition. The defensibility lies in the proprietary AI models and the data generated through their discovery process. However, the 'UX quality' is not yet evident, and the 'platform potential' is speculative at this early stage. The actual efficacy and scalability of the AI agents will be key to long-term defensibility.
The company is very early stage, with no explicit mention of revenue or paying users. Being part of Y Combinator is a positive signal and indicates early investor interest. The press coverage is positive but standard for YC launches. Partnerships and significant customer adoption are likely still in the future. The lack of concrete traction metrics at this point is the primary limiting factor for a higher score. [Boost +2: Tier-1 VC: accel]
News
Matforge, founded by Akash Ramdas and Advaith Sridhar, is among the Y Combinator Spring 2026 startups, focusing on building AI scientists to discover new semiconductor materials.
Matforge is developing AI scientists to discover novel semiconductor materials, aiming to shorten the lengthy research and development cycle and potentially revive Moore's Law.
Matforge is building AI agents to accelerate the discovery of new materials for the semiconductor industry, aiming to reduce the typical 10+ year lab work to months.
This page provides an analysis of Matforge's funding, indicating it is a company focused on AI for material discovery.
MatForge offers a platform that accelerates materials R&D by remembering past experiments, designing optimal DOE matrices, and using AI to suggest next steps.
Matforge, a Y Combinator Spring 2026 startup, is developing AI scientists to discover semiconductor materials, compressing the discovery timeline from over 10 years to months using AI agents.
Matforge is building AI scientists to accelerate the discovery of new materials for the semiconductor industry, aiming to find 10x better alternatives in months instead of years.
Quick Info
- Batch
- Spring 2026
- Team Size
- 2
- Location
- San Francisco, CA, USA
- Founders
- 2
- Scraped
- 4/10/2026