Chip Industry Week In Review


Arm uncorked its first internally developed CPU chip this week, aimed squarely at the agentic AI data center market. Arm CEO Rene Haas (pictured) emphasized the CPU's power efficiency and performance/watt compared to other AI processor architectures. "We are obsessed with efficiency, and if you think about one of the biggest appeals that Arm has had over the years, it is power profile," he ... » read more

Chip Industry Week In Review


War impacts The Iran War's toll on the chip industry is widening. Over 95% of Taiwan's energy is imported, causing the country to secure alternative sources. Korea is also heavily dependent on energy imports from the Middle East. Shortages of key materials are cropping up everywhere. Helium from Qatar, the second largest producer behind the U.S., is constrained by hostilities in the Per... » read more

Advanced Packaging Limits Come Into Focus


Key Takeaways: Packaging is now a performance variable. Substrate, bonding, and process sequence determine what can be built at scale. Warpage underlies most advanced packaging failures and gets harder to control as package sizes grow. Every proposed solution, such as glass, panel processing, and backside power, solves one problem while creating another. Moore's Law has shif... » read more

Replay‑based Validation as a Scalable Methodology for Chiplet‑based Systems (Intel, Synopsys)


A new technical paper, "ODIN-Based CPU-GPU Architecture with Replay-Driven Simulation and Emulation," was published by researchers at Intel, Nvidia and Synopsys. Abstract "Integration of CPU and GPU technologies is a key enabler for modern AI and graphics workloads, combining control-oriented processing with massive parallel compute capability. As systems evolve toward chiplet-based archite... » read more

Chip Industry Week In Review


Disruptions caused by the Iran conflict have taken about one third of the global helium supply off the market, an essential gas for semiconductor manufacturing, reports the World Economic Forum. Other potential impacts for the chip industry include bromine and other chemical shortages, logistical disruptions, and higher energy prices incurred by fabs in Asia. Top Deals IBM and Lam R... » read more

Scale Up, Scale Out Get a New Partner


Key Takeaways: Three AI data center scaling strategies are scale-up, scale-out, and scale-across. Scale-up is within a rack; scale-out is between racks; scale-across is between data centers. Each of the three uses a different interconnect strategy to optimize either latency or jitter. As today’s data center workloads — especially for AI and HPC — outgrow the physical, ... » read more

Chip Industry Week In Review


Think tank IAPS' report on AI integrity attacks contends that advanced AI systems must be protected from hidden tampering, backdoors, or unauthorized changes that could alter their behavior or outputs, especially when AI adoption is scaling rapidly, with over 60% of the federal workforce now using AI every day. Geopolitics The U.S. government has drafted new export rules that may give W... » read more

Chip Industry Week In Review


Big Deals and Fundings Rapidus secured US$1.7B in a new funding round from the Japanese government and the private sector to ramp 2nm production by next year. Open AI announced a $110B in new funding, with $30B from Nvidia, $30B from Softbank and $50B from Amazon. In a $100B multi-year deal, Meta will power its AI infrastructure with up to 6GW of AMD's GPUs. SambaNova and Intel ar... » read more

Verifying Scale-Up And Scale-Out In Data Centers


Semiconductor Engineering sat down to discuss challenges and solutions for data center build-out and build-up with Gordon Allan, Siemens EDA director of verification IP; Rishi Chugh, vice president of product marketing for network switching at Marvell; Saravanan Kalinagasamy, senior director of ASIC design and validation at Astera Labs; and Jalaj Gupta, product engineering lead at Siemens EDA. ... » read more

Chiplets 2026: Where Are We Today?


Jim Handy of Objective Analysis and Jawad Nasrullah from Palo Alto Electron kicked off last week's Chiplet Summit with predictions about where the chiplet market is headed and why chiplets are needed to accelerate AI. Handy noted that in the 1990s, multi-chip modules (MCMs) led to mid-'90s multi-chip packages (MCPs), and then progressed to NAND flash stacking, stacked die, big chips (e.g., X... » read more

← Older posts Newer posts →