Inference Acceleration: Follow The Memory


Much has been written about the computational complexity of inference acceleration: very large matrix multiplies for fully-connected layers and huge numbers of 3x3 convolutions across megapixel images, both of which require many thousands of MACs (multiplier-accumulators) to achieve high throughput for models like ResNet-50 and YOLOv3. The other side of the coin is managing the movement of d... » read more

The Winograd Transformation


Cheng Wang, senior vice president of engineering at Flex Logix, explains how the Winograd Transformation applies to convolutional neural networks. https://youtu.be/E7QJUby9x-I » read more

Embedded FPGA Basics


You do not need know about FPGAs to integrate reconfigurable RTL into your SoC: our software maps your RTL into our EFLX array for you.  But if you are curious, read on. FPGAs are field programmable gate arrays. They offer a different kind of programmability from processors.  Processors are sequential while FPGAs enable massive parallelism.  A processor has one adder, one multiplier—an ... » read more

Blog Review: Feb. 13


UltraSoc’s Rupert Baines notes that awareness is rapidly increasing for the value of embedded insights and system data. Cadence’s Paul McLellan reviews the highlights of the upcoming SPIE conference in San Jose. Synopsys’ Taylor Armerding discusses the key findings of a report on cybersecurity practices for the automotive industry, including what components of vehicles pose the hi... » read more

Week in Review: IoT, Security, Auto


Internet of Things McKinsey & Company identified 10 top trends in the Internet of Things. They include: IoT is a business opportunity, not just a tech opportunity; disciplined execution across multiple use cases is the path to value; and IoT is gradually enabling more subscription business models, but consumers are resistant. Louis Columbus of IQMS provides some IoT data points and id... » read more

Use Inference Benchmarks Similar To Your Application


If an Inference IP supplier or Inference Accelerator Chip supplier offers a benchmark, it is probably ResNet-50. As a result, it might seem logical to use ResNet-50 to compare inference offerings. If you plan to use ResNet-50 it would be; but if your target application model is significantly different from Resnet-50 it could lead you to pick an inference offering that is not best for you. ... » read more

Benchmarks For The Edge


Geoff Tate, CEO of Flex Logix, talks about benchmarking in edge devices, particularly for convolutional neural networks. https://youtu.be/-beVEpKAM4M » read more

EDA Grabs Bigger Slice Of Chip Market


EDA revenues have been a fairly constant percentage of semiconductor revenues, but that may change in 2019. With new customers creating demand, and some traditional customers shifting focus from advanced nodes, the various branches of the EDA tool industry may be where sticky technical problems are solved. IC manufacturing, packaging and development tools all are finding new ways to handle t... » read more

Week In Review: Design, Low Power


M&A Ansys will acquire Helic, a provider of electromagnetic crosstalk analysis and signoff tools. Founded in 2000, Helic's tools included pre- and post-LVS EM modeling, inductor synthesis and modeling, and analysis of crosstalk risk. The company's technology will be integrated into a solution for on-chip, 3D integrated circuit and chip-package-system electromagnetics and noise analysis. Th... » read more

Week in Review: IoT, Security, Auto


Internet of Things Tony Franklin, Intel’s general manager for Internet of Things Segments, is interviewed by Lorin Fries on how the chipmaker is helping to develop smart farming applications. “We focus primarily on high-performance computer technologies, as well as communication technologies, which have great applicability for food systems. We work closely with a broad ecosystem of partner... » read more

← Older posts