Keep pulling the thread on Nagesh Gupta.
According to a 2024 study by Wilson Research and Siemens, only 14% of chips were successful after the first tape-out.
A 2024 study by Wilson Research and Siemens found that 80% of embedded systems projects require rework.
Multiple studies suggest that if a product is delayed by more than 6 months, 33% of its lifetime revenue is lost.
A re-tapeout for a chip using current technology costs tens of millions of dollars and takes at least 6 months to complete.
Lip-Bu Tan has set a goal for Intel that the B0 tape-out must be the final version ready for production.
Nagesh Gupta founded Lambda AI approximately two years ago to address the multi-year, iterative nature of chip design.
Rework performed at later stages of a project costs about 100 times more than rework at earlier stages.
The lack of horizontal optimization across product definition, SoC design, hardware design, and embedded software is a major cause of project delays in the semiconductor industry.
Lambda AI is natively agentic and has developed several complete platforms based on multiple agents performing various tasks.
Customers are increasingly pushing to take their first chip tape-out (A0) directly to production release.
Lambda AI has released a product called LambdaSpectra that automates the generation of production-quality technical documents.
Lambda AI's technology stack includes an LLM orchestration layer and a platform layer built on top of a base LLM like Claude.