Memory Architectures In AI: One Size Doesn’t Fit All


In the world of regular computing, we are used to certain ways of architecting for memory access to meet latency, bandwidth and power goals. These have evolved over many years to give us the multiple layers of caching and hardware cache-coherency management schemes which are now so familiar. Machine learning (ML) has introduced new complications in this area for multiple reasons. AI/ML chips ca... » read more

Smart Cities, Challenging Issues


Smart cities are coming. Not everything will be connected, and not everything will be connected at once. Still, governments around the globe are beginning to tap into a world of connected devices and sensors for reasons ranging from cheaper lighting to less traffic, lower crime, and improved air quality. Smart cities encompass all manner of usage models and equipment — parking meters, traf... » read more

Seeing The Future Of Vision


Vision systems have evolved from cameras that enable robots to “see” on a factory floor to a safety-critical element of the heterogeneous systems guiding autonomous vehicles, as well as other applications that call for parallel processing technology to quickly recognize objects, people, and the surrounding environment. Automotive electronics and mobile devices currently dominate embedded... » read more

Plugging Holes In Machine Learning


The number of companies using machine learning is accelerating, but so far there are no tools to validate, verify and debug these systems. That presents a problem for the chipmakers and systems companies that increasingly rely on machine learning to optimize their technology because, at least for now, it creates the potential for errors that are extremely difficult to trace and fix. At the s... » read more

The Week In Review: Manufacturing


Chipmakers GlobalFoundries has rolled out its next-generation FD-SOI technology. The new 12nm FD-SOI process is called 12FDX. It is designed for a range of applications, from mobile computing and 5G connectivity to artificial intelligence and autonomous vehicles. "Some applications require the unsurpassed performance of finFET transistors, but the vast majority of connected devices need high l... » read more

The Week In Review: Design


M&A Intel will acquire Movidius, adding the company's low-power vision processing unit to its growing computer vision efforts that include a depth-sensing camera and machine learning projects. At the same time, Intel is shedding cybersecurity unit McAfee (acquired in 2011 for $7.7 billion and re-named Intel Security in 2014). Intel will retain a 49% stake in the business with the 51% rem... » read more

What’s Missing From Machine Learning


Machine learning is everywhere. It's being used to optimize complex chips, balance power and performance inside of data centers, program robots, and to keep expensive electronics updated and operating. What's less obvious, though, is there are no commercially available tools to validate, verify and debug these systems once machines evolve beyond the final specification. The expectation is th... » read more

The Week In Review: Design


M&A The FTC has given the go-ahead to ON Semiconductor's acquisition of Fairchild Semiconductor. As part of the requirements, ON Semiconductor had to divest its planar insulated gate bipolar transistor business, which will be sold to Littelfuse. (Littelfuse will also pick up the transient voltage suppression diode and switching thyristor product lines for a combined $104 million in cash.... » read more

The Week In Review: Design/IoT


Tools Synopsys unveiled a new custom design solution targeting FinFET layout, introducing visually-assisted routing automation, a built-in design rule checking engine, templates to apply previous layout decisions to new designs, and IC Compiler integration. TSMC certified the new tool for 10nm and 7nm FinFET process technologies. It has also been adopted by STMicroelectronics, GSI Technology... » read more