Author's Latest Posts


Week In Review: Design, Low Power


Tools & IP Cadence unveiled deep neural-network accelerator (DNA) AI processor IP, Tensilica DNA 100, targeted at on-device neural network inference applications. The processor is scalable from 0.5 TMAC (Tera multiply-accumulate) to 12 TMACs, or 100s of TMACs with multiple processors stacked, and the company claims it delivers up to 4.7X better performance and up to 2.3X more performance p... » read more

Blog Review: Sept. 19


Applied Materials' David Thompson shares the new DARPA program that is focused on using correlated electrons to develop a new type of switch with quantum effects, potentially leading to unprecedented switching speeds. Mentor's Joe Hupcey III argues that for the most effective formal analysis, assertions should be as simple as possible and shares some tips on decomposing big assertions. Ca... » read more

Power/Performance Bits: Sept. 18


Etching photovoltaics Researchers at Michigan Technological University and Aalto University found a way to reduce production costs of black silicon solar cells by more than 10%. The first prototype modules have been manufactured on an industrial production line. Typically, the silicon used for solar cells is etched to reduce reflected light, although some light is still lost. Nano-texturing... » read more

Week In Review: Design, Low Power


M&A Intel acquired NetSpeed Systems, a network-on-a-chip and interconnect fabric IP and tool provider. Founded in 2011, the San Jose-based company recently put a focus on interconnects designed with AI applications in mind. Intel has cast the acquisition as a way to tie a number of its other technologies together. The team will join Intel's Silicon Engineering Group. Intel has been a NetSp... » read more

Blog Review: Sept. 12


Cadence's Paul McLellan checks out the impact the Meltdown, Spectre, and Foreshadow vulnerabilities will have on future processor design with an overview of speculative execution and why it's important to current architectures. Mentor's Matthew Ballance suggests some ways to find existing information and descriptions that can be used to jump-start the creation of portable stimulus models. ... » read more

Power/Performance Bits: Sept. 11


Non-toxic photoluminescent nanoparticles Researchers from Osaka University developed a way to improve display technologies using non-toxic light-emitting nanoparticles. In trying to replace cadmium and other toxic materials used in quantum dots, scientists have turned to non-toxic nanoparticles that emit light in an efficient manner by creating I–III–VI semiconductors, such as silver in... » read more

Week In Review: Design, Low Power


Cadence teamed up with nine PCB manufacturing partners on an ecosystem to provide easier access to partners' technology files to improve PCB manufacturability. The program, DesignTrue DFM, allows for automated import of a manufacturer's latest DFM rules as well as rule checking in real time as part of the PCB layout process. The nine initial PCB manufactures supporting the program are: Bay Area... » read more

Blog Review: Sept. 5


Synopsys' Taylor Armerding looks at a case of exploding costs – up to $17 million – when the city of Atlanta, Georgia fell victim to the SamSam ransomware, plus the lessons other cities can take to improve their security. Cadence's Paul McLellan traces how a landmark moment in object recognition, the ImageNet database, has spurred increasingly better object recognition algorithms for alm... » read more

Power/Performance Bits: Sept. 4


Preventing battery fires Researchers from Oak Ridge National Laboratory and the University of Rochester developed a method to prevent lithium-ion batteries from catching on fire when damaged. "In a lithium-ion battery, a thin piece of plastic separates the two electrodes," said Gabriel Veith, a research lead at ORNL. "If the battery is damaged and the plastic layer fails, the electrodes can... » read more

Week In Review: Design, Low Power


Synopsys unveiled a new formal app, Regression Mode Accelerator, which uses machine learning algorithms to speed up formal property verification, as well as better convergence of formal proofs for subsequent runs. According to Synopsys, the app also allows for significant saving of compute resources in nightly regressions. Hitachi used OneSpin Solutions' 360 EC product family to verify vCOSS... » read more

← Older posts