Toward domain-specific architectures; floating point in embedded; Sonics acquisition and the interconnect market.
Cadence’s Paul McLellan argues that rapid improvements in the performance of general-purpose computing led to a lack of innovation in domain-specific architectures, but as scaling slows, they’re entering a new golden age.
In a video, Mentor’s Colin Walls takes a look at the use of floating point in an embedded application and some of the pitfalls associated with it.
Synopsys’ Taylor Armerding follows up on European privacy regulation GDPR and finds while the number of fines levied is not huge and neither so far are the penalties, the quiet start is not expected to last.
Semico Research’s Rich Wawrzyniak digs into Facebook’s acquisition of Sonics and what it means for the future direction of the interconnect IP market.
Arm’s Kobus Marneweck contends that the nature of the IoT means taking physical security seriously from the beginning of a design is of vital importance and explains the different types of attacks to prepare for.
In a video, VLSI Research’s Dan Hutcheson learns the rules Mark Bohr of Intel sets out for how to innovate at scale based on what worked, and didn’t, during his career.
China-based blogger Stewart Randall argues that in spite of its best efforts, China’s quest for semiconductor self-sufficiency has a lot of hurdles in the way, and looks at the fab ecosystem and design tools to understand why.
ANSYS’ Shawn Wasserman checks out a new way to build MEMS using a layered method from startup Integra Devices that allows devices to be built of materials incompatible in a standard manufacturing process.
SEMI’s Maria Vetrano chats with Sean Ding of Alibaba about the role of sensors pushing data processing to the edge and the right approach to AI in IoT devices.
For more good reading, check out the highlighted blogs from last week’s Low Power-High Performance newsletter:
Editor In Chief Ed Sperling contends that tolerances are going to be very tight at the next nodes, where research is just beginning.
Rambus’ Frank Ferro explains how PCB materials and vias can address insertion loss and crosstalk in automotive GDDR6.
Cadence’s Rohit Kapur shows the different DFT technologies and when to insert them into a design.
Mentor’s Harry Foster questions whether verification techniques are keeping up with FPGAs’ growing complexity.
Synopsys’ Vadhiraj Sankaranarayanan demonstrates how unique architectural features tailor DRAM for specific applications.
 |
 |
 |
 |
 |
 |
 |
 |
 |  |
Leave a Reply