The Rise Of AI Co-Processors


Figuring out the best kinds of processors to use for different AI workloads is a challenge. AI algorithms are undergoing rapid and frequent changes, and the workloads tied to them can vary by data type, by user, and sometimes because of software/firmware updates. On top of that, AI computations tend to require much higher utilization rates than traditional computing, and that will only become m... » read more

Benefits And Challenges Of Using Chiplets


The move to chiplets opens the door to more features than can be packed into a reticle-sized SoC. That potentially means more processing power, simpler designs, and higher yields. But it's not as simple as swapping LEGO blocks into a chassis. Ashley Stevens, director of product management and marketing at Arteris, talks with Semiconductor Engineering about the challenges of using coherent versu... » read more

Silicon Lifecycle Management


How chips are used is changing, and so are the requirements. In the past, markets were largely segmented by application, which determined how chips were designed. High-performance processors went into notebook computers, low-power chipsets were deployed in mobile devices, and complex SoCs and advanced packages were used in data centers. But with the spread of AI everywhere, traditional segmenta... » read more

Virtual Metrology In Semiconductor Manufacturing


Fourth in a seven-part series: Virtual metrology may never be 100% perfect because of the almost unlimited number of changes in a fab tools and the unique chip and wafer designs they're being used to process. But there are places where virtual metrology does make sense. Jon Herlocker, vice president and general manager of software analytics at Cohu, talks about why virtual metrology will never ... » read more

Virtual Twins: Layers Of Challenges


Virtual twins can provide deep insights into complex systems at any point in time, but creating them requires integrating a stack of abstractions that don't naturally go together. One abstraction may be mechanical, another electrical, and the data used to create those abstraction layers needs to be fused together logically and updated over time. David Fried, corporate vice president at Lam Rese... » read more

Inside Chips Podcast: Data Movement In The AI Age


AI is all about data movement — lots of it. The key is to move data as little as possible, and when it is moved, to do it efficiently, securely, and blindingly fast. Semiconductor Engineering talks with Arteris CEO Charlie Janac in this one-on-one discussion about the impact of AI on networks on chip and what will change going forward. To listen to the podcast, click here. » read more

The Evolution of DRAM


DRAM has been around since 1966, but today it's still the same basic 1T 1C bit cell architecture. Yet changes are coming as DRAM is called upon to store and retrieve more data faster. Steve Woo, distinguished inventor and fellow at Rambus, talks about how DRAM works, why there are different flavors, the impact of cooling new solutions in denser configurations, and ongoing issues involving the s... » read more

Using AI For Fault Detection And Classification In Manufacturing


Third in a seven-part series: Classic fault detection and classification has some classic problems. It's reactive, time-consuming to set up, and any product change involves significant man-hours. Even then, it still misses a lot of problems, which result in scrap. This is where machine learning can excel, because it can sift through huge amounts of data from thousands of sensors and find outlie... » read more

Challenges In Stacking HBM


AI data centers are pushing for higher density in high-bandwidth memory. Today, the maximum number of layers that can be stacked is 8, but that increases to as many as 24 layers by 2030. The big challenge will be in the interconnects, and making sure the microbumps align. At 16 layers, the bump pitch will be less than 10 microns, and the dies will be thinner. Damon Tsai, head of product marketi... » read more

Preparing For The Quantum Computing Age


Within a decade, quantum computers will be able to break virtually any encryption algorithm in use today. What used to be science fiction is on its way to becoming a commercial reality. Once that happens, quantum computers will be able to crack in minutes what was supposed to be unbreakable for more than a century using the most powerful computers available. Erik Wood, senior director of crypto... » read more

← Older posts Newer posts →