Podcast EP339: Unique Scalable, Power-Efficient AI Technology from EdgeCortix with Dr. Sakya Dasgupta

Summary

Daniel is joined by Dr. Sakya Dasgupta, CEO and founder of EdgeCortix. He is a seasoned technologist who has been at the forefront of artificial intelligence systems for the past two decades. His experience spans public research institutes such as The Max Planck Society and RIKEN, as well as leading corporations like Microsoft… Read More

Key takeaways

• Edge Cortex's Sakura 2 delivers 60 TOPS at under 8 watts as a co-processor, enabling large language models like Llama 3 (8B parameters) to run on power-constrained edge systems without replacing existing host processors.
• Sakura 2 passed NASA radiation tolerance testing as a commercial off-the-shelf (COTS) device—achieving 10x higher radiation tolerance than competing solutions without special packaging—unlocking cost-effective AI compute for LEO, geosynchronous, and lunar missions.
• Edge Cortex secured ~$50M in non-dilutive Japanese government funding and became the first Japanese company awarded a direct contract with the U.S. Defense Innovation Unit, validating its dual-use aerospace and defense positioning.

Why this matters

As edge AI deployments in robotics, aerospace, and defense demand high-performance inference under severe power and size constraints, Edge Cortex's co-designed hardware-software approach and radiation-hardened COTS validation represent a credible alternative to GPU-centric architectures for mission-critical applications.

Entities

Show notes

Daniel is joined by Dr. Sakya Dasgupta, CEO and founder of EdgeCortix. He is a seasoned technologist who has been at the forefront of artificial intelligence systems for the past two decades. His experience spans public research institutes such as The Max Planck Society and RIKEN, as well as leading corporations like Microsoft… Read More

Themes

• artificial-intelligence