Blog Review: Nov. 18

rss · SemiEngineering 2026-11-18T08:01:49Z en
Cadence’s Rajesh Pamujula considers how to establish trust when sensitive data, proprietary models, cryptographic keys, and confidential workloads increasingly traverse PCIe links. Synopsys’ Neeraj Paliwal argues that extreme co-design, where silicon, interface IP, memory, cooling, and power are designed together as one system, is needed for efficient racks that are capable of running increasingly demanding AI workloads. Siemens’ Kevin Rinebold and Andras Vass-Varnai examine the evolving semiconductor design landscape shaped by the shift from monolithic scaling to heterogeneous architectures and consider how to navigate the complexities of AI-driven 3D-IC and chiplet technologies. Keysight’s Linas Dauksa provides a primer on the technologies that power AI data centers, how they handle training and inference workloads, and why they’ve become the backbone of modern AI. Arm’s Michael Gamble checks out how to enable gaze-based interactions in lightweight smart glasses by running fast, sensor-dependent AI on-device while larger workloads stay where more compute is available. SEMI’s Sitong He chats with Patrick Vandenameele of imec about the trends reshaping the semiconductor industry, why collaboration matters more than ever, and the role Europe can play in global competitiveness. Plus, check out blogs featured in the latest Low Power-High Performance newsletter: Rambus’ Kushal Koolwal examines why longer contexts, higher concurrency, multimodal models, and AI agents will continue changing where memory pressure appears. Siemens EDA’s Jim Shiely shows how to use evidence from silicon to improve the tools and decisions for next-gen hardware. Synopsys’ Andrew Appleby observes that traditional voltage scaling is no longer sufficient to deliver real-time performance within tight energy and thermal limits. Cadence’s Joe Chen demonstrates how to right-size the memory hierarchy around the device’s data streams, reuse patterns, latency requirements, and operating constraints. Arm’s Siyu Mou explains the rationale behind 8-bit floating-point data formats, their ecosystem momentum, and hardware support. The post Blog Review: Nov. 18 appeared first on Semiconductor Engineering.
PCIe and trust; rack-scale co-design; heterogeneous scaling; AI data center primer; on-device balancing; EU semi trends. Cadence’s Rajesh Pamujula considers how to establish trust when sensitive data, proprietary models, cryptographic keys, and confidential workloads increasingly traverse PCIe links. Synopsys’ Neeraj Paliwal argues that extreme co-design, where silicon, interface IP, memory, cooling, and power are designed together as one system, is needed for efficient racks that are capable of running increasingly demanding AI workloads. Siemens’ Kevin Rinebold and Andras Vass-Varnai examine the evolving semiconductor design landscape shaped by the shift from monolithic scaling to heterogeneous architectures and consider how to navigate the complexities of AI-driven 3D-IC and chiplet technologies. Keysight’s Linas Dauksa provides a primer on the technologies that power AI data centers, how they handle training and inference workloads, and why they’ve become the backbone of modern AI. Arm’s Michael Gamble checks out how to enable gaze-based interactions in lightweight smart glasses by running fast, sensor-dependent AI on-device while larger workloads stay where more compute is available. SEMI’s Sitong He chats with Patrick Vandenameele of imec about the trends reshaping the semiconductor industry, why collaboration matters more than ever, and the role Europe can play in global competitiveness. Plus, check out blogs featured in the latest Low Power-High Performance newsletter: Rambus’ Kushal Koolwal examines why longer contexts, higher concurrency, multimodal models, and AI agents will continue changing where memory pressure appears. Siemens EDA’s Jim Shiely shows how to use evidence from silicon to improve the tools and decisions for next-gen hardware. Synopsys’ Andrew Appleby observes that traditional voltage scaling is no longer sufficient to deliver real-time performance within tight energy and thermal limits. Cadence’s Joe Chen demonstrates how to right-size the memory hierarchy around the device’s data streams, reuse patterns, latency requirements, and operating constraints. Arm’s Siyu Mou explains the rationale behind 8-bit floating-point data formats, their ecosystem momentum, and hardware support. Jesse Allen   (all posts) Jesse Allen is the Knowledge Center administrator and a senior editor at Semiconductor Engineering.
Highlight