In its ninth year, Design World’s LEAP Awards recognize the engineering innovations shaping the future of design and technology across a range of categories. Each year, the awards bring together leading OEM design engineers and academics to form an independent judging panel that evaluates the submissions and selects the most outstanding solutions. The program reflects the innovation and expertise driving progress across the engineering community.
This year’s winners in the Connectivity category highlight advances in wireless networking, AI infrastructure, and high-speed data transmission, showcasing technologies that address the growing demands of an increasingly connected world.
Congratulations to the LEAP Awards for Connectivity winners, who are profiled here:
GOLD
Morse Micro
MM8108 Edge AI Connectivity Platform
Morse Micro’s MM8108 Edge AI Connectivity Platform is redefining wireless connectivity for the AI era by extending secure, IP-based networking far beyond the traditional reach of Wi-Fi. Built on Wi-Fi HaLow technology, the platform addresses one of the greatest barriers to Edge AI deployment: delivering reliable, long-range connectivity to the cameras, sensors, industrial equipment, energy systems, and smart buildings that make up intelligent infrastructure. Rather than forcing engineers to choose between high-throughput wireless with limited coverage and long-range networks with minimal bandwidth, the MM8108 enables reliable communication across campuses, factories, warehouses, utilities, agricultural operations, and commercial facilities.
Supporting kilometer-scale coverage, enterprise-grade WPA3 security, and more than 8,000 connected devices per access point, the MM8108 gives engineers the flexibility to deploy AI-enabled systems across expansive environments while simplifying network architectures. Its long-range capabilities allow cameras to monitor areas beyond building walls, utility assets to be tracked across distributed networks, and industrial equipment to support predictive maintenance throughout large production facilities. By combining standards-based networking, low power consumption, and extended range in a single embedded connectivity platform, Morse Micro provides the communications foundation for IoT 2.0 and the next generation of Edge AI. As artificial intelligence moves beyond the cloud and into the physical world, the MM8108 is helping engineers connect intelligent systems wherever they can deliver value, unlocking new possibilities for scalable, secure, and efficient wireless infrastructure.
SILVER
Credo
OmniConnect Weaver
Credo’s OmniConnect Weaver redefines AI memory connectivity by addressing one of the most pressing challenges in modern AI infrastructure: the memory wall. As AI models grow in complexity and inference workloads demand greater memory capacity and bandwidth, traditional packaging architectures can limit system performance and scalability. OmniConnect Weaver is the industry’s first memory fanout gearbox designed for scalable, high-bandwidth AI inference, enabling AI accelerators and other compute processors (xPUs) to access significantly larger memory pools while improving computing efficiency. By extending memory connectivity beyond the physical boundaries of the compute substrate, Weaver enables disaggregated memory configurations that overcome conventional on-package and on-substrate limitations.
Combining ultra-efficient 112G VSR SerDes, a lightweight AXI framer, and Credo’s advanced interconnect architecture, OmniConnect Weaver increases I/O density by up to 10 times, supporting up to 5.8 TB of LPDDR5X memory and 16 TB/s of bandwidth. As a memory connectivity chiplet, it functions as a high-speed memory fanout gearbox between AI accelerators and LPDDR5X memory, expanding the reach, capacity, and bandwidth of memory connections without requiring traditional packaging constraints to dictate system design. This architecture gives engineers greater flexibility to scale memory resources to meet the demands of increasingly sophisticated AI inference workloads while maintaining the performance and efficiency required for large-scale deployment. By enabling denser memory configurations and more scalable chip-to-chip connectivity, OmniConnect Weaver provides a new approach to AI system architecture, helping unlock the next generation of high-performance, memory-intensive AI infrastructure.
SILVER
Marvell
1.6T Optical DSP Platform Portfolio
Marvell’s 1.6T Optical DSP Platform Portfolio is advancing optical connectivity for the AI era, addressing the growing demands of AI infrastructure as hyperscalers build data centers containing hundreds of thousands and ultimately millions of GPUs and AI accelerators. As connectivity becomes a critical factor in system performance, efficiency, and scalability, traditional optical architectures designed for conventional cloud networking are increasingly challenged to deliver the bandwidth density, resiliency, and power efficiency required by next-generation AI networks. Marvell addresses these demands with a comprehensive family of AI-optimized optical digital signal processors designed specifically for 1.6T connectivity. The portfolio includes four specialized DSPs, Ara™ T, Ara™ X, Petra™, and Aquila™ M, each optimized for distinct applications spanning intra-data center, campus-scale, and scale-across AI infrastructure.
The platform introduces a range of innovations, including transmit-retimed optics (TRO), advanced optical link resiliency, 3nm gearbox technology, and integrated MACsec security. These technologies enable cloud providers and equipment manufacturers to transition from 800G infrastructure to 1.6T optical networks with increased bandwidth, lower latency, reduced power consumption, and greater deployment flexibility. Beyond its individual DSPs, the portfolio integrates with Marvell’s broader connectivity technologies, including silicon photonics, coherent optics, PAM4 DSPs, Ethernet PHYs, advanced SerDes, and switching solutions, to provide a comprehensive foundation for next-generation AI networking. By combining high-speed optical connectivity with improved resiliency, security, and energy efficiency, Marvell’s 1.6T Optical DSP Platform Portfolio empowers engineers to build larger, faster, and more efficient AI networks, helping meet the escalating connectivity demands of large-scale AI infrastructure.