Toward Software-Defined Vehicles


Speed is everything when it comes to designing automotive electronics, but not in the usual way. In the past, product cycles often lasted five to seven years, from initial design to implementation inside of vehicles. That no longer works as vehicles adopt more electronic features to replace mechanical ones, and as competition heats up over the latest features and nearly instantaneous over-the-a... » read more

Changes In Formal Verification


For the better part of two decades, formal verification was considered too difficult to use in many designs and too slow for anything but narrow bug hunting. Much has changed recently. Ashish Darbari, CEO of Axiomise, explains why formal is now essential for finding deadlocks, security holes, and Xprop issues in mission-critical, safety-critical, and AI designs, and how that will apply to chipl... » read more

Promises And Pitfalls Of SoC Restructuring


As chips become more complex and increasingly heterogeneous, it's becoming more difficult to keep track of different methodologies, tools, and blend data from different sources to create a chip. Tim Schneider, staff application engineer at Arteris, explains why IP-XACT has become so critical, why it took so long to gain a solid foothold in chip design, and how the new IP-XACT standard interface... » read more

Making Adaptive Test Work Better


One of the big challenges for IC test is making sense of mountains of data, a direct result of more features being packed onto a single die, or multiple chiplets being assembled into an advanced package. Collecting all that data through various agents and building models on the tester no longer makes sense for a couple reasons — there is too much data, and there are multiple customers using t... » read more

MCU Changes At The Edge


Microcontrollers are becoming a key platform for processing machine learning at the edge due to two significant changes. First, they now can include multiple cores, including some for high performance and others for low power, as well as other specialized processing elements such as neural network accelerators. Second, machine learning algorithms have been pruned to the point where inferencing ... » read more

Electromigration And IR Drop At Advanced Nodes


Manufacturing chips at 3nm and below is a challenge, but it's only part of the problem. Designing chips that can be manufactured and will actually work is potentially an even bigger problem. There is more data to sift through for place-and-route, less margin to pad a design, and there are more physical effects to contend with as transistors get taller, density increases, and chips age. Jeff Wil... » read more

Adapting To Evolving IC Requirements


As chip designs become increasingly heterogeneous and domain-specific, packing a device with one-size-fits-all chips or chiplets doesn't make sense. The key is rightsizing different components based on real workloads, so they don't waste power when there is too little utilization of logic, and so they don't struggle to complete tasks because they are undersized. Jayson Bethurem, vice president ... » read more

Sensor Fusion Challenges In Automotive


The number of sensors in automobiles is growing rapidly alongside new safety features and increasing levels of autonomy. The challenge is integrating them in a way that makes sense, because these sensors are optimized for different types of data, sometimes with different resolution requirements even for the same type of data, and frequently with very different latency, power consumption, and re... » read more

Overlay Optimization In Advanced IC Substrates


Overlay is becoming a significant problem in the manufacturing of semiconductors, especially in the world of advanced packaging substrates — think panels — the larger the area, the greater the potential for distortion due to warpage. Solving this issue requires more accurate models, better communication through feed forward/feed back throughout the flow, and real-time analytics that are bak... » read more

Secure Movement Of Data In Test


Historically, test data flowed out of the tester and was loaded into a file. But with heterogeneous integration, including chiplets and IP from multiple vendors, test data is now being streamed across the manufacturing floor where it can be used to make real-time decisions. Eli Roth, product manager for smart manufacturing at Teradyne, talks with Semiconductor Engineering about challenges in da... » read more

← Older posts