Blog Review: May 20

Cadence’s Siddh Virani demonstrates how to import and integrate foreign language logic into PSS on both Target and Solve platforms, opening possibilities for code reuse and cross-language collaboration.

Synopsys’ Sumit Vishwakarma finds that AI model training and inference workloads are forcing the industry to rethink not only how much compute fits in a rack, but how servers are architected from end to end, resulting in a complete transformation of computing infrastructure.

Siemens’ Emily Yan compares HBM3e and HBM4, identifies key design challenges sich as signal integrity and thermal management, and considers the projected timeframes for adoption.

Intel Foundry’s Ravi V. Mahajan argues that while heterogeneous integration is emerging as the most practical path to sustaining AI growth, more detailed roadmaps are needed to guide how the industry stacks, connects, powers, and cools tomorrow’s chips.

Arm’s Chris Armstrong provides an overview of why optimized sparse linear algebra functions are important and introduces an open-source project for sparse linear algebra on Arm, along with the latest updates such as support for supernodal triangular solves.

Keysight’s Muhammad Umar Khan explains why end-to-end electrical-optical-electrical simulation is key for building reliable, high-performance Ethernet systems.

The ESD Alliance’s Bob Smith chats with Joe Kwan of Siemens EDA about what’s driving design and manufacturing collaboration and what can be done to improve design success.

And don’t miss the blogs featured in the latest Low Power-High Performance newsletter:

Expedera’s Athish Rahul Rao explains why peak TOPS is becoming a weaker proxy for actual edge performance.

Rambus’ Piero Bianco digs into preserving LPDDR’s energy efficiency advantages while restoring the modularity required for server systems.

Quadric’s Mike Leonard details an emerging AI architecture for embedded autonomy that improves edge efficiency.

Arm’s Jade Alglave introduces an experimental AI chatbot that acts as a guide to the Arm architecture, providing quick answers to complex technical questions.

Cadence’s Veena Parthan compares the accuracy of two meshing workflows when dealing with complex blade geometries.

Siemens EDA’s Carey Robertson explains how AI can speed sign-off while increasing confidence that a design will function as expected.