I/O Design Challenges Grow In AI Data Centers And HPC Clusters


Key Takeaways: A designer’s choice of I/O connectors and interconnect protocols can be the difference between a massively profitable AI chip and a flop. I/O tradeoffs impact airflow, cooling, rack design, power coming into the rack, and other critical aspects of HPC chip design. Reliability is paramount, so standards must be followed, and I/Os need redundant pins. Other innovations... » read more

Wafer-Scale vs. Chiplets: The New War? Part 2


In Part 1, we looked at the innovations underpinning the Cerebras WSE-3 and why its most significant breakthrough is the elimination of data movement overhead at the architectural level, not better yield management or thermal engineering. Cerebras’ on-wafer fabric is a viable answer to the question being asked by the entire industry: how do you move data fast enough that compute stops wait... » read more

Designing Chips That Can Explain Themselves


Key Takeaways: On-die telemetry gives architects a path to replace worst-case design margin with measured silicon behavior, improving PPA without compromising resilience. As monitor density and control-loop speed increase, observability must be architected hierarchically across local hardware response, on-die processing, and fleet-level learning. The real payoff is architectural: str... » read more

Chip Industry Week In Review


Notable deals Cadence and Intel Foundry inked a multi-year agreement to advance design technology co-optimization and create PDKs for Intel Foundry's 14A process. Nvidia and SK hynix announced a multi-year partnership to co-develop memory technology for AI infrastructure and physical AI. Teradyne unveiled an integrated test cell solution with TEL that supports known-good device scree... » read more

Chip Industry Week In Review


Computex in Taiwan: Arm and Nvidia introduced an AI PC platform, RTX Spark, with an Arm-based Grace CPU, Blackwell RTX GPU, and unified memory. Cadence announced a fully autonomous virtual agentic AI design engineer, enabling customers to run dynamic simulations in automated workflows. Intel launched Xeon 6+, its first data-center CPU built on Intel Foundry's 18A process. The company... » read more

Swapping Out Chiplets: I/Os Vs. Compute


Key Takeaways: Companies can save time and money by swapping out a compute, memory, or I/O chiplet to gain technology improvements, while keeping the other dies stable. Chip architects may choose to keep their I/Os stable and swap out compute to move from a 5nm process node to 3nm to achieve performance and power improvements, or swap out memory from LPDDR5X to LPDDR6. Swapping out... » read more

Observability Is Essential For Modern Silicon


Experts At The Table: In-silicon observability — also known as on-die or on-chip visibility — is becoming increasingly important for managing the performance, reliability, and security of today’s high-performance systems. Semiconductor Engineering sat down to discuss this with Andy Nightingale, vice president of product management and marketing at Arteris; Nandan Nayampally, chief commerc... » read more

Wafer-Scale vs. Chiplets: The New War? Part 1


Cerebras’ IPO is a meaningful moment for the semiconductor industry — and not just for the financial implications. Their confidence in their opening price reflects something the industry has effectively acknowledged: incremental chip scaling can no longer keep pace with what AI infrastructure demands. Radical approaches are earning serious consideration and serious capital. Cerebras... » read more

Overcoming Bottlenecks In Data Movement


AI is all about data. There is more data to process, store, and move, and more tradeoffs required to do that efficiently and with enough flexibility to handle changes in future workloads. Nandan Nayampally, chief commercial officer at Baya Systems, talks about networks on chip and networks across chip, what the choke points are for data movement, and where and when data coherency makes sense. » read more

Confusion Grows With More Interconnect Options And Tradeoffs


Key Takeaways: Designers are frequently evaluating 5 or more different interconnects in a single system, each with a distinct purpose. While chip-to-chip (PCIe) and die-to-die (UCIe, BoW) technologies seem to be solving a similar problem, in practice they bring different challenges. PCIe, CXL, NVLink, and UALink are all active in the hyperscaler space, but Ethernet-based technologies... » read more

← Older posts