Rambus Captures 3 Top Spots in Semi Engineering’s 2018 Top Tech Talks

This entry was posted on Friday, January 25th, 2019.

Kudos go to Rambus for scoring three big wins in Semiconductor Engineering’s 2018 Top Tech Talks. Semi Engineering staffer Linda Christensen reported on its site’s videos that had the most traffic and why.

The three Rambus videos were presented by Frank Ferro, Sr. Director, Product Marketing for Memory and Interfaces Division, Steve Woo, Vice President, Systems and Solutions, Office of the CTO, and Rambus Chief Scientist, Craig Hampel.

According to Christensen, “Traffic numbers on stories give a snapshot of the most current trends, but with videos those trends are even more apparent because of the time invested in watching those videos.”

Semiconductor Engineering categorizes the 2018 Top Tech Talks videos into automotive, memory, system-level design, low-power high performance, manufacturing & process slots, and FPGAs, AI & communication slots.

Frank Ferro got the nod for his video HBM Vs. GDDR6. He talked about memory bottlenecks and focused on key points why GDDR6and high-bandwidth memory are gaining greater popularity in several markets.

He talked about the demands of such applications as automotive and AI and noted that processor speeds are outstripping the memory bandwidth available to the processor.  Plus, he added, SoC designers are looking for ways of how to open up that memory bandwidth with existing architectures.

Also, in the Memory category, Steve Woo spoke about on how memory architectures are shifting in Data-Driven Design.  He cited memory hierarchies and how they are changing as the amount of data continues to grow.  In part, Steve said architecture is all about balance.  “So, you’re trying to make sure you’re meeting the needs of whatever the application is.  We’re starting to see some evolution in terms of what’s going on in the processing, as well as in the memory system that stores the data.  For certain types of systems, you’re seeing novel types of processing come to market.   One such area is machine learning where people are building specialized chips to do the processing for neural networks,” he added.

In the Low-Power High Performance area, Craig Hampel explained to his video audience what will drive the next big performance gains after Moore’s Law from the data center to the edge in Huge Performance Gains Ahead.  In part, he said, “At Rambus, we’re looking at changing the operating environment of the computer with data centers being as centralized and as a bigger part of these businesses of those large conglomerates.”