Marvell (Israel)

This paper introduces a unified algorithmic design flow for low-noise amplifiers (LNAs), spanning specifications to layout and integrating topology, architecture, circuit, and electromagnetic (EM) design in one framework. The algorithm supports designer input from classical to unconventional architectures, enhancing interpretability, debugging, and usability. We demonstrate two LNAs covering 24 to 150GHz with state-of-the-art performance, including novel architectures uncovered by the algorithm.

Abstract Flip chip ball grid array (FCBGA) is one of the most broadly used package types in different types of electronic packages. In a FCBGA package, the die/ silicon is flipped and attached to substrate with help of copper pillars. The material property coefficient of thermal expansion (CTE) mismatch between die/ silicon and substrate materials like core, dielectric material results in package thermal warpage and may lead to thermo-mechanical related failures in electronic packages. CTE can be defined as the rate at which a material expands with change in the temperature. Material expansion increases with a higher CTE. The CTE of die/ silicon is lower than the overall CTE of organic substrate. During reflow, as an electronic package cools down it warps in a convex or crying shape which indicates positive package warpage, and as the electronic package heats up it warps in a concave or smiling shape which indicates negative package warpage. Induced thermal warpage during the electronic package reflow assembly process is an important factor affecting the package assembly and thermo-mechanical reliability of FCBGA packages; therefore, precise package thermal warpage prediction is one of the most significant parameters for the electronic package design processes. In this study, the effect of different electronic package parameters (core material CTE, core thickness, package type) of the FCBGA package on its warpage are assessed by using finite element analysis (FEA) mechanical simulation methodology. A bare die FCBGA package having substrate core material with lower coefficient of thermal expansion (CTE) shows lower package warpage as the mismatch between die/ silicon and substrate core material CTE becomes dominant factor to drive the package warpage. To understand the package warpage trend, warpage is evaluated at room temperature & high/ reflow temperature. The package warpage trend is reversed for larger FCBGA packages with stiffener ring. Substrate core material with higher CTE shows lower package warpage as CTE mismatch between substrate core material and stiffener ring material (steel/copper) CTE becomes the dominant factor to drive the package warpage. The counter intuitive package warpage data from mechanical simulation is compared with shadow-moire test data and it shows good correlation and validates the simulations findings. Based on this work, it is recommended that a thorough investigation for large size packages be conducted before choosing packaging materials. This work shall provide new insights for package designers to optimize thermal warpage and stress during package initial design phase.

Secure key transport and storage between Hardware Security Modules (HSMs) is a critical requirement in cryptographic systems, particularly in offline or air-gapped environments where network connectivity cannot be assumed. This paper introduces KeyBlob, a standardized format designed to address HSM key management challenges, including offline security, multi-key transfers, rich metadata handling, and key origin authenticity. KeyBlob uniquely combines support for multiple wrapping standards with the encapsulation of multiple keys in a single, cohesive format. It enhances security through flexible key wrapping, a Cryptographic Message Syntax (CMS) SignedData structure for origin authenticity, optional padding to obscure key sizes, and per-key MACs for integrity. Practical features include compressed attributes for efficiency, attribute encryption for flexible access control, PKCS\#8 compliance for private key management, and protocol-specific key wrapping for HSM clusters. An optimized request structure and multi-key support via an ASN. 1 SEQUENCE OF WrappedKey enables efficient offline HSM-to-HSM key exchange. Comparative analysis with standards like TR-31, TR-34, and the Key Management Interoperability Protocol (KMIP) highlights KeyBlob's distinctive advantages: unparalleled flexibility in key type and wrapping support, robust CMS-rooted security, and a design tailored for offline scenarios. Comprehensive evaluation and design objectives comparison demonstrate KeyBlob's practicality as a cutting-edge solution for contemporary HSM key management challenges.

Resource-constrained edge devices can offload compute-intensive, latency-sensitive video inference tasks to nearby edge cloud servers. In such deployments, network delays typically have a limited impact on the overall end-to-end response time. However, data copying and processing delays within the edge server's inference pipeline-spanning the NIC, CPU, memory, and GPU-are becoming a key performance bottleneck. To address this, we present FR-Infer, a GPU-centric Fast and Reliable Inference pipeline for edge-offloaded video inference. FR-Infer bypasses the host CPU by enabling the NIC to transfer request data directly into GPU memory (NIC-GPU), significantly reducing delays in the pipeline. Two key challenges tackled in FR-Infer include: (1) ensuring reliable and ordered delivery over unreliable edge-access networks, and (2) supporting efficient data pre-processing to prepare video frames for DNN inference. We prototype FR-Infer on a server equipped with a Bluefield-2 DPU and NVIDIA A5000 GPU, and show that it reduces inference response time by <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$21-49 \%$</tex> compared to traditional NIC-CPU-GPU pipelines.

We review EEPN effects across different spectral ranges. For coherent pluggables, 1/f noise -induced timing drift can be dominant. We show the bursty feature of resulting performance degradation. We discuss recent advancements in mitigation techniques. ©2025 The Author(s)

This study examines the link between spend rewards and vouchers and how they influence consumer behavior. Spend rewards and vouchers are strong drivers that initiate purchasing behavior, alter the spending habits of consumers, and build reward-based consumer loyalty. This study aims to determine the link between spend rewards and vouchers and examines how these promotional tools affect the purchasing behavior of consumers in e-commerce and online shopping. The quantitative design uses a questionnaire to collect data from internet consumers. The questionnaire collects data on consumers’ spending, the consumers’ attitudes towards vouchers and rewards, and usage frequency of the incentives. Random sampling was used to select participants for the study. Statistical analysis involves determining the percentage distribution of the responses to identify patterns and relationships. The findings indicate a high positive relationship between spend rewards and voucher availability and higher consumer spending. A large proportion of the respondents state that these rewards have an impact on their purchasing behavior, resulting in greater participation and repeat business. These findings imply that providing spend rewards and vouchers is a sound approach to increasing customer retention and driving sales. In conclusion, the research verifies that spend rewards and vouchers has a positive effect on consumer behavior by making purchases more frequent and building brand loyalty. The findings give confidence that the adoption of such incentive schemes can yield positive results for companies in the e-commerce sector.

Every year Kiwa PVEL’s Photovoltaic (PV) Module Reliability Scorecard highlights the technical advancements and limitations of commercially available solar modules. Modules included in the Scorecard are subjected to a series of extended reliability and performance tests, known as the Product Qualification Program (PQP), and the aggregated data and analysis is published at scorecard.pvel.com. This has become the industry’s leading publication on PV module reliability and offers readers a wide range of key insights. In June 2025, Kiwa PVEL will release the 11<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sup> edition of this annual report. While the Scorecard often showcases the advances made in PV manufacturing, it also reveals concerning trends. For example, the 2024 Scorecard reported that two-thirds of participating manufacturers and 41% of bill of materials (BOMs) experienced at least one test failure. Specific PQP tests, like thermal cycling (TC) and damp heat (DH), showed significant failure rates. This paper will offer a deep dive into the latest findings from the 2025 Scorecard, which promise to be as impactful as those included in previous editions.

In inertial confinement fusion, pure deuterium-tritium (DT) is usually used as a fusion fuel. In their paper, S. Y. Guskov et al. [Plasma Phys. Rep. 37, 1020 (2011)] instead propose using low-Z compounds that contain DT and are non-cryogenic at room temperature. They suggest that these fuels can be ignited for ρDTR≥0.35gcm−2 and kTe≥14keV, i.e., parameters that are more stringent but still in the same order of magnitude as those for DT. In deriving these results, Guskov et al. assume that ionic and electronic temperatures are equal and consider only electronic stopping power. Here, we show that at temperatures greater than 10 keV, ionic stopping power is not negligible compared to the electronic one. We demonstrate that this necessarily leads to higher ionic than electronic temperatures. Both factors facilitate ignition, showing that non-cryogenic DT compounds are more versatile than previously known. In addition, we find that heavy beryllium borohydride ignites more easily than heavy beryllium hydride, the best-performing fuel found by Guskov et al. Our results are based on an analytical model that incorporates a detailed stopping power analysis, as well as on numerical simulations using an improved version of the community hydro code MULTI-IFE. Alleviating the constraints and costs of cryogenic technology and the fact that non-cryogenic DT fuels are solids at room temperature opens up new design options for fusion targets with Q&amp;gt;100. The discussion presented here generalizes the analysis of fuels for energy production.