A case for why the AI infrastructure buildout is justified even if the bubble in the stock prices is real — built deliberately on the most conservative market-size estimate available, not the most exciting one.
Read the full piece →Most Intel vs. TSMC discussions miss the most important engineering question — Intel's High-NA lithography bet and what it would mean if it works.
How Cerebras revived an idea that failed catastrophically in 1980, and what that says about who actually survives a hardware arms race.
Memory and processor architecture for anyone who wants to go past the surface-level explanation of why AI hardware looks the way it does.
Most explanations of AI either oversimplify or overwhelm. This one tries to do neither — using linear algebra as the lens into what a neural network actually is, mathematically.
A historical perspective on all five generations of wireless technology and what is genuinely new in 5G — written before most people had seen a 5G phone in the wild.
The three technology islands 5G introduces — enhanced mobile broadband, ultra-reliable low latency, and massive machine-type communications. Each a platform for a new wave of services.
Kumar Venkatramani
President & CEO, Silicon Ideas Inc.
I've spent the past 40+ years at the intersection of hardware and software — building tools, IP, and service portfolios for everyone from early-stage startups to Fortune 500 companies in the semiconductor world. Along the way I've filed 14 patents in design methodology and system-level design and published in IEEE and ACM conferences. I hold an MS in Electrical Engineering from Columbia University (1983) and a BE Honours in Electronics and Communication from NIT Trichy, India (1981).
I write these essays to go deeper on the technology questions that don't get enough serious treatment — the engineering decisions, the architectural tradeoffs, and occasionally the economics underneath them. Written for people who want to look under the hood, not just read the headline.