Why More CPUs Are Needed For Agentic AI


The shift from generative AI to agentic AI will significantly increase the amount of compute power needed in data centers. Queries to search for and analyze data from multiple sources will be performed simultaneously by agents and without human intervention, rather than a single request from a live person. Jeff Defilippi, senior director of product management at Arm, talks about the impact of r... » read more

Blog Review: Apr. 15


Cadence's Wilson Kobalkar shares why eUSB2‑V2 represents a major evolutionary step for the USB 2.0 ecosystem, including how it achieves multi‑gigabit HSx operation and why symmetric/asymmetric modes unlock new design possibilities. Synopsys' Akanksha Soni explains the difference between metal-oxide-metal, metal-insulator-metal, and metal-oxide-semiconductor capacitors, identifying the ad... » read more

Startup Funding: Q1 2026


The new year started off with a bang for private semiconductor companies, with 18 garnering mega funding rounds exceeding $100 million, and two, Rapidus and Cerebras, reaching the $1 billion mark. Predictably, the vast majority of those are either designing chips primarily for AI inference workloads or attempting to overcome bandwidth limitations by improving interconnects from the chip level t... » read more

A New Era For Co-Processing


Key Takeaways: There is no single processor capable of executing everything efficiently, meaning that multiple processors are required. Maximum efficiency is gained by minimizing the movement of data. Architects must maximize efficiency for today's workloads, while also adding enough flexibility to handle tomorrow's. New processor architectures are rapidly evolving thanks to... » read more

Rethinking Robotics Reinforcement Learning: A Practical Humanoid Training Workflow


Reinforcement learning (RL) for robotics is often associated with large GPU clusters, distributed infrastructure, and x86-based development environments. Training a humanoid robot with high-fidelity simulation is a resource-intensive workflow that runs in the data center. What if that workflow could run on a single workstation? In this blog post, we explore a complete robotics pipeline bu... » read more

Fast Isn’t Fast Enough: Redefining Metrics for Edge AI


Key Takeaways: Edge AI performance is about low latency and power efficiency, not peak TOPS. Memory bandwidth and data movement now limit edge AI more than compute. Successful edge AI requires balanced hardware, software, and fast model updates. Experts At The Table: Today’s chip architect must contend with multiple factors when architecting AI processors for fast and effi... » read more

Blog Review: Apr. 8


Cadence's Shyam Sharma highlights new capabilities in LPDDR6, including metadata built into the data packets, rowhammer mitigations, DVFS with support for three operating voltage rails, and new efficiency modes. Synopsys' Akanksha Soni points to multiphysics simulation as a key element of ensuring automotive IC designs meet ISO 26262 requirements. Siemens' John McMillan suggests a simulat... » read more

Chip Industry Week In Review


Deals IBM and Arm are collaborating on a new dual‑architecture hardware aimed at enterprise AI and data-intensive workloads, using virtualization to boost reliability, security, scalability, and software compatibility. The goal, according to an IBM spokesperson, is to deliver side-by-side deployments of S390x-Linux and Arm-Linux virtual machines in a single kernel-based hypervisor. Nv... » read more

Blog Review: April 1


Siemens EDA's Harry Foster considers why first-silicon success is continuing to decline even though tools are capable of handling much larger design sizes and identifies how increasingly complex interactions between components cause traditional verification assumptions to break down. Synopsys' Eldo N Baby explores dynamic voltage drop analysis, including how to bring in switching scenario in... » read more

Challenges In Scaling Chips To 2nm And Below


Key Takeaways Scaling to 2nm and below continues due to power improvements per watt, but progress is much more challenging and costly. Solutions to problems often create other problems due to less margin for tradeoffs, often requiring larger interposers, more chiplets, and more complex packages. New levels of precision are required throughout the design-through-manufacturing flow, re... » read more

← Older posts Newer posts →