11 Ways To Reduce AI Energy Consumption


As the machine-learning industry evolves, the focus has expanded from merely solving the problem to solving the problem better. “Better” often has meant accuracy or speed, but as data-center energy budgets explode and machine learning moves to the edge, energy consumption has taken its place alongside accuracy and speed as a critical issue. There are a number of approaches to neural netw... » read more

HBM Takes On A Much Bigger Role


High-bandwidth memory is getting faster and showing up in more designs, but this stacked DRAM technology may play a much bigger role as a gateway for both chiplet-based SoCs and true 3D designs. HBM increasingly is being viewed as a way of pushing heterogenous distributed processing to a completely different level. Once viewed as an expensive technology that only could be utilized in the hig... » read more

Is AI Good Or Bad For The Planet?


Will artificial intelligence save or sink planet earth? We’re surrounded by AI. When you use the internet, take a photo, use predictive text, or watch TV, you are interacting with AI. And we are still in the early stages of this revolution in technology and our lives. But AI can require large amounts of power. Researchers have documented the astounding amount of power required to train ... » read more

Mapping Heat Across A System


Thermal issues are becoming more difficult to resolve as chip features get smaller and systems get faster and more complex. They now require the integration of technologies from both the design and manufacturing flows, making design for power and heat a much broader problem. This is evident with the evolution of a smart phone. Phones sold 10 years ago were very different devices. Functionali... » read more

Blog Review: May 12


Cadence's Claire Ying points to major changes in PCIe 6.0 as PAM4 signaling replaces NRZ to help double bandwidth, Forward Error Correction helps maintain data integrity, and various improvements are made to power consumption. Synopsys' Samantha Beaumont argues that automotive sensors are a major potential attack point and addresses some of the key areas of sensor vulnerability and the chall... » read more

Blog Review: May 5


Arm's William Wang considers how to increase the performance and programmability of persistent applications through using battery to protect the on-chip volatile cache hierarchy. Cadence's Paul McLellan finds that ransomware is getting more sophisticated, and more difficult to eradicate and defend against, with potentially life-threatening consequences. Synopsys' Jonathan Knudsen digs int... » read more

Steep Spike For Chip Complexity And Unknowns


Cramming more and different kinds of processors and memories onto a die or into a package is causing the number of unknowns and the complexity of those designs to skyrocket. There are good reasons for combining all of these different devices into an SoC or advanced package. They increase functionality and can offer big improvements in performance and power that are no longer available just b... » read more

Week In Review: Design, Low Power


IP, FPGA, Tools Arm released new details on its new Neoverse N2 and Neoverse V1 platforms. A range of companies announced they will be using the platforms, including Marvell and SiPearl. Aimed at server and HPC workloads, Neoverse V1 uses wider and deeper pipelines compared to the N1 and supports a 2x256bit wide vector unit executing the Scalable Vector Extension (SVE) instructions with sup... » read more

Week In Review: Auto, Security, Pervasive Computing


Pervasive computing — IoT, edge, cloud, data center, and back Arm shared new features and some design wins for its Neoverse V1 and N2 platforms, which it introduced in September 2020. Neoverse V1 is optimized for high-performance computing. Arm has licensing wins from India’s Ministry of Electronics and Information Technology (MeitY) for its exascale HPC project; and European exascale comp... » read more

Blog Review: April 28


Arm's Tiago Azevedo explains why it's important to measure uncertainty when using object detection, especially in critical applications such as automotive, and introduces a architecture that can do so while balancing complexity and efficiency. Cadence's Paul McLellan digs into the ISO 21434 standard for addressing and managing cybersecurity risks in vehicles and why it's a good sign for secu... » read more

← Older posts Newer posts →