In this episode of Baya TechThreads, Raja Koduri and Dr. Shailesh Kumar dive into the scaling challenges of the AI era and how engineers are solving them. From chiplets to software-defined fabrics, they break down how simplified, modular architectures are enabling massive, heterogeneous compute. If you’re building for performance, scalability, or the trillion-agent future, this is essential listening.
In this episode of Baya TechThreads, Raja Koduri and Dr. Shailesh Kumar dive into the scaling challenges of the AI era and how engineers are solving them. From chiplets to software-defined fabrics, they break down how simplified, modular architectures are enabling massive, heterogeneous compute. If you’re building for performance, scalability, or the trillion-agent future, this is essential listening.
Join Laura Milstein and Kevin Albert as they chat with Nandan Nayampally, Chief Commercial Officer at Baya, about the future of AI and intelligent computing. Nandan reveals how Baya tackles the critical issue of data movement, a major bottleneck in AI innovation. He offers insights into the pivotal role of hardware and data flow in next-gen AI systems. Learn about the chip landscape, the rise of hybrid cloud-edge AI, and the impact of tariffs on the semiconductor supply chain.
Baya Systems will be speaking at the upcoming GSA TECH Summit in San Jose, California on July 1st. This premier industry event brings together semiconductor leaders to explore the technologies and trends shaping the future of compute. We’re excited to contribute to the conversation with our perspective on scaling AI, overcoming data movement challenges, and designing infrastructure for the next generation of intelligent systems. If you’re attending, we’d love to connect. Lunch is on us!
From silicon to system design, our latest blogs dive into the hidden bottlenecks and bold ideas shaping the future of AI infrastructure. Here’s what we’ve been working on.
Explore how Baya Systems and Imagination Technologies used CacheStudio™ at RISC-V Silicon Valley to profile cache behavior across CPU, GPU, and hybrid setups. The blog reveals eye-opening findings on how cache sizing and coherence impact data movement, memory reuse, and system throughput—all crucial for architects building truly efficient heterogeneous platforms.
Dr. Shailesh Kumar challenges the notion that bespoke fabrics always win. This post lays out why many in-house designs struggle with scalability, mounted costs, and delayed timelines—while showing how Baya’s modular, software-defined fabric IP has helped teams cut engineering cycles and pivot midstream with confidence.
News Spotlight: Scaling Up and Scaling Out: Reimagining Silicon Switching for the AI Era
We are hiring! With rapid growth, we’re expanding our engineering (software and hardware) and commercial teams. We encourage you to explore opportunities and join the team.
Copyright (C) 2025 Baya Systems. All rights reserved.
You are receiving this email because you opted in via our website.
Our mailing address is:
Baya Systems · 5201 Great America Pkwy · Santa Clara, CA 95054-1122 · USA