Items Tagged with 'ML'

Since woven glass PCB substrates are anisotropic, EDA design and modeling software hoping to advance AI and ML algorithms should have provisions to model anisotropic material, especially via transitions. In this article, Bert Simonovich discusses the importance of having an awareness of the test method used by CCL suppliers for accurate modeling and simulation. Simonovich covers how the use of out-of-plane Dk_z values instead of in-plane Dk_xy values for via modeling can cause misleading simulation results, which may result in reduced margins and potential compliance test failures when the design is built and tested. 

Keysight Technologies, Inc. has entered the AI and ML infrastructure ecosystem with the introduction of the Keysight AI Data Center Test Platform, designed to accelerate innovation in AI/ML network validation and optimization.

The New Trident 5-X12 doubles bandwidth, reduces power by 25%, and adds neural network to enable next-generation telemetry, security, and traffic engineering.

As AI and ML technologies continue to evolve and become more prevalent in various industries, the demand for faster and more efficient data transfer methods will undoubtedly grow. In this guest blog, VIAVI reviews how PCIe 6.0 and CXL 3.x, with their advanced capabilities and perfect alignment, are already paving the way for rapid innovation in AI/ML use cases. 

Molex has announced the results of a global reliability survey that reveals the challenges facing system architects and design engineers for hardware, including devices, when balancing growing expectations for reliability with ever-increasing product complexity, diminishing time for testing, and constant cost and manufacturing constraints.

At this year’s Brooklyn 6G Summit, visitors can experience the first demonstration of how a neural receiver paired with a learned custom modulation performs in a 5G NR UL MIMO scenario, from Rohde & Schwarz and NVIDIA.

Automotive applications present new challenges to high-speed serial technology. Asymmetric, multi-gigabit signaling between sensors, processors, and displays in the unique noise environments of both electric and internal combustion engine vehicles create new problems for signal and power integrity engineers. This paper introduces the signal impairments required for receiver testing in the emerging automotive standards like ASA, MIPI's A-PHY, Automotive Ethernet, and more. Standards specify different sources of noise in different ways, some in the form of time evolutions, others as spectra. This paper focuses on techniques for generating and calibrating each noise source while describing advanced de-embedding techniques and addressing test equipment limitations.