Marathon is Bungie’s new survival extraction FPS game that blends bold design with fast paced loot focused gameplay. Whether you’re fighting against the UESC forces, or other players – you’ll need a fast and stable connection to succeed.
In this blog we’ll show you how to fix all connection issues in Marathon. Whether you’re suffering from high ping, packet loss or an unstable connection, we have a fix for you in this blog.
The first thing to check when experiencing high ping or connection issues in Marathon is the games server status. If the servers are experiencing problems or become overloaded at peak times, your connection quality may suffer or your may experience issue logging into the game.
The only official way to check Marathon’s server status is through the official Bungie server status X (Twitter) Account, which you can find here.
Unfortunately, there is no direct solution for server issues as it’s up to the game developers to fix them. But knowing they are the cause of your connection issues will save you time troubleshooting your own network.
Using a wired connection instead of Wi-Fi will drastically improve your connection quality in Marathon. Whilst Wi-Fi is convenient for getting your devices online, it has some shortcomings for online gaming.
The main issues come from signal strength and interference, which can lead to packet loss, rubber banding and disconnections. In a high stakes game like Marathon where packet loss or disconnections can cause you to lose value loot and time, you need to make sure your connection is as stable as possible.
Switching to a wired ethernet connection is one of the quickest ways to get a better connection in Marathon. Ethernet offers a faster, more stable and lower latency connection – which is ideal for fast paced looting and gunfights. It also does not suffer from signal strength or interference, like Wi-Fi.
The only drawback to ethernet is whether its practical for you to run a cable between your router and gaming device. If this is an issue for your setup, you may want to look into Powerline Adapters. These let you carry an ethernet connection between devices using your home existing electrical wiring.
The Netduma R3 gaming router is the ultimate tool to combat connection issues in Marathon. Whether you’re suffering from high ping, an unstable connection or packet loss – the R3 has a fix.
In the following sections we’ll break down the gaming features of the R3 that will eliminate each connection issue so you can get back to scavenging around Tau Ceti IV.
The Geo-Filter feature allows you to control what game servers you connect to, instead of leaving it all up to a matchmaking algorithm.
The game server you connect to is the biggest determining factor of your ping in Marathon. The further away the server is, the higher your ping, this mean you’ll always want to connect to your nearest servers.
The Geo-Filter allows you to do this by creating a customisable filter around your location which forces connection to servers inside the radius, whilst blocking connection to servers outside of it.
This feature also gives you insight into your connection quality, which you can’t normally see due to Marathon having no connection metric settings. The side bar of the Geo-Filter will show you the server you’re connected to, your in-game ping and how stable your connection is.
If you want to learn more about the Geo-Filter and its more advanced settings, check out our Ultimate Geo-Filter guide below.
There is nothing worse in an intense gunfight than your ping spiking, especially when your loot is on the line. Steady Ping fixes this issue by stabilising your connection and eliminating jitter. This helps the game better sync to the game server you’re connected to, which results in extremely responsive inputs and gameplay.
Once Steady Ping is enabled, it will automatically start to stabilise your connection as you load into a game. If you want more control, or just want to tinker with settings, there is an Expert Mode available which let’s you set your ideal ping to stabilise to.
Packet loss is a major problem in fast paced games like Marathon, as dropped packets can completely ruin your chances of an exfil. This is an extremely problem common on busy networks, with other devices and applications fighting over your networks bandwidth.
The R3’s SmartBOOST feature allows you to prioritise gaming and eliminates bufferbloat, so all applications on your network will run smoother. By ensuring big downloads or streaming movies doesn’t interrupt your gaming, you’ll always be playing on a fast and stable connection.
If you want to see SmartBOOST optimise a busy home network, check our our live demo video below.
SAN LUIS OBISPO, Calif., March 10, 2026
Gateworks Corporation, a leader in industrial embedded technology, announced the expansion of its Catalina Single Board Computer (SBC) family. Engineered around the high-performance NXP® Semiconductors i.MX 95 applications processor, the Catalina platform is designed to bring transformative Edge AI, machine vision and autonomous capabilities to the harshest industrial environments.
NXP will feature the i.MX 95 processor with live application demonstrations at Embedded World, March 10-12th in Nuremberg, Germany.
The Catalina SBC is optimized for three primary pillars of modern industrial technology:
The Catalina series marks a deepening of the collaboration between Gateworks and NXP. As an NXP® Semiconductors Gold Partner, Gateworks ensures seamless integration of the NXP i.MX 95 SoC’s advanced features, including the integrated eIQ® Neutron NPU and ISP.
At NXP, we focus on enabling secure, scalable edge intelligence across industrial and embedded markets. Gateworks brings proven engineering expertise, high-quality USA manufacturing, and a ruggedized platform approach to the i.MX 95 applications processor. Together, we will continue to empower customers to deploy sophisticated AI applications with confidence and speed.
Robert Thompson, Director, Smart Connected Edge Ecosystem, NXP Semiconductors
The Catalina family introduces a “no-compromise” architecture for high-bandwidth industrial applications. Available in the compact GW9200 (90 x 100mm) and GW9400 (170 x 100mm), the platform combines extreme networking with field-proven reliability:
Recognizing the rapid evolution of AI models, Gateworks is designing modular expansion paths based on NXP’s AI technology. This allows customers to scale AI performance as requirements grow without redesigning their base compute architecture.
Catalina provides the ‘industrial-grade’ backbone that edge AI and intelligent automation have been waiting for. By pairing NXP’s cutting-edge silicon and AI acceleration with our US-based manufacturing and design expertise, we are helping our customers deploy smarter, more resilient systems at scale.
Kelly Peralta, VP of Sales and Business Development, Gateworks Corporation
NXP will showcase application-driven demos focusing on industrial vision and robotics-ready architectures powered by the i.MX 95 SoC. Meet NXP at Embedded World, Booth 4A-222
Gateworks’ i.MX 95-based Development Kits and production quantities are available now through distribution and directly through Gateworks.
For more information or to get started with an evaluation, view the Catalina GW9200 Product Page.
As organizational AI strategies mature, we are seeing a clear shift from broad-spectrum, generic AI platforms, such as Microsoft Copilot or Google Gemini, toward purpose-built, customized intelligence trained on proprietary data. While early AI models relied on bolt-on external sources, enterprises are now rapidly pivoting towards more advanced, built-in, fit-for-purpose AI, designed for specific use cases, such as predictive maintenance or fraud detection. These models are trained and fine-tuned on relevant domain data, integrated into core systems and built to support decision making and deliver measurable business outcomes.
“Embedded” or “native” AI is justified because native intelligence promises more nuanced performances and superior outcomes, and greater security, while making better, long-term economic sense.
Let’s take a deeper look at why the shift from generic external AI to embedded AI is a pivot in the right direction.
General-purpose AI, by its very nature, relies on generic, external data. In the absence of customized data, specific compliance frameworks, or regulatory obligations, the deep, detailed and granular insights required fail to materialize. An algorithm that is not tailored to meet a specific business objective cannot deliver real value by producing results that are broad, superficial, and lacking nuance. Additionally, external AI is disconnected from the operational protocols, infrastructure nuances, real-time constraints, or an understanding of the hierarchy of specific service provider systems.
This context deficit has the potential to impact outcomes adversely, as it cannot provide the basis for intelligent, informed decision-making that any company would require.
All AI models need data to train their algorithms. While external AI uses synthetic or generic data, built-in AI ingests real telemetry data to build native intelligence. This knowledge helps embedded AI learn to recognize operational patterns, identify behavioral anomalies, accurately predict outcomes, and even avoid failures; all tailored to the company for which it was built.
Unlike external AI, embedded AI models learn from real-world events, user behavior, and system dynamics. They develop a high level of domain fluency that is invaluable and almost impossible for external solutions to replicate.
The responses generated reflect actual systems without translation layers, which are basic to external AI. Moreover, the answers are not simulated examples. They are instead grounded in real infrastructure.
Edge computing is a distributed computing model that processes data at or near the source rather than on centralized cloud servers. The physical sources could be IoT devices, local servers, or regional data centers. Processing on the edge has undeniable advantages. It reduces latency, improves speed, and enables real-time decision-making.
Edge-deployed intelligence eliminates latency associated with data transmission traffic delays caused by network hops, multiple authentication layers, and queue processing. With embedded AI, when network detects unusual traffic patterns, systems can respond locally in microseconds, without round-trip communication with external systems.
This proves to be a decisive advantage during times of security incidents or performance degradation. Native AI intercepts threats and reroutes traffic before incidents occur, ensuring users experience no disruption.
Every microsecond saved adds to millions of decisions daily. The cumulative impact is faster threat mitigation, reduced downtime, and superior user experience.
In enterprises, data sovereignty and security containment are top priorities across the board. Embedded solutions ensure sensitive telemetry data stays within the security perimeter. Operational metrics, user behavior patterns, and infrastructure health data remain secure and contained. Embedded AI is an architectural choice designed to safeguard sensitive data and prevent security breaches.
External AI carries inherent risks. Every API call transmits intelligence to third-party systems. Even with encryption and contractual safeguards, exposing systems to expanded vulnerability points increases compliance exposure.
Operational data represents the most accurate picture of the service provider’s systems. Generic AI models often carry noise, bias, and inaccuracies. Built-in intelligence trains on verified, controlled, and curated information, ensuring greater data sovereignty and relevance.
Another clear advantage of native/embedded AI is the transparent cost structures. The upfront investment in development and integration supports predictable scaling alongside infrastructure growth. Operational expenses remain proportional to actual usage.
External AI imposes perpetual variable costs. API charges, data transfer fees, and integration overhead compounds indefinitely. Higher usage triggers billing surprises. Vendor pricing changes can shock and derail budget planning.
In network optimization, these costs can be high. External solutions charge per query, causing costs to escalate, often unpredictably, as network operations scale up. In contrast, embedded intelligence operates on predictable infrastructure costs, processing billions of routing decisions without per-transaction fees.
Built-in AI integrates seamlessly with existing operational frameworks. Monitoring, alerting, and orchestration tools already speak the same language. Updates happen through established pipelines. Troubleshooting uses familiar, pre-set methodologies.
On the other hand, external solutions fragment the service provider’s operational landscape. Each integration requires custom connections, security configurations, authentication protocols, and error handling, adding complexity to the process. Managing multiple vendor relationships, SLAs, and support channels becomes an additional task. Overhead rises as more external tools are added, teams waste time switching between platforms, documentation piles up, and knowledge silos are created.
Native or embedded intelligence goes beyond technical superiority. It is architectural sovereignty. It’s the control over the capabilities that differentiates a service provider’s operations from its competitors.
External AI treats intelligence as a commodity. Every competitor gains identical capabilities. There is no uniqueness. And operational advantages diminish when everyone accesses the same third-party source.
Network optimization makes this dynamic clear. Embedded AI learns specific traffic patterns, partner and peer relationships, and capacity constraints. It generates proprietary insights that external models, trained on aggregated industry data, cannot match.
Built-in intelligence improves over time. As infrastructure evolves, native AI adapts organically. External solutions force adaptation to vendor updates and changes, not the other way around.
The choice between embedded and external AI ultimately determines whether intelligence becomes a competitive moat or a rented commodity. For organizations serious about operational excellence, native intelligence isn’t just preferable. It’s essential.
[Taipei, Taiwan – March 17, 2026] – AsiaRF, a global leader in wireless connectivity solutions, is proud to announce it has achieved Rank 243 in the “High-Growth Companies Asia-Pacific 2026” list, compiled by the Financial Times and Statista.
This prestigious ranking identifies the top 500 companies in the Asia-Pacific region that achieved the highest compound annual growth rate (CAGR) in revenue between 2021 and 2024. AsiaRF’s inclusion reflects its exceptional performance and rapid expansion in the competitive wireless communication market.
Driving the Future of Connectivity In recent years, AsiaRF has reached significant milestones through the development and mass production of cutting-edge technologies, including Wi-Fi 7 (802.11be) modules and Wi-Fi HaLow (802.11ah)solutions. By providing robust, high-performance hardware for IoT, industrial automation, and enterprise networking, AsiaRF has solidified its position as a trusted partner for B2B clients worldwide.
“We are honored to be recognized by the Financial Times as one of the fastest-growing companies in the region,” said by Paul Lai, the CEO of AsiaRF. “This achievement is a testament to our team’s dedication to innovation and our commitment to delivering world-class wireless solutions. As we look forward, we remain focused on scaling our impact and helping our customers navigate the evolving landscape of global connectivity.”
About the FT High-Growth Companies Asia-Pacific 2026 The FT High-Growth Companies Asia-Pacific 2026 is a joint initiative by the Financial Times and Statista. It ranks the top 500 companies from 13 countries and territories across the region based on revenue growth. The full rankings and detailed analysis are available on FT.com.
? Check out FT ranking: High-growth companies Asia-Pacific 2026 full rank
About AsiaRF Founded in 1996, AsiaRF is a premier manufacturer of wireless communication solutions. Specializing in high-performance Wi-Fi modules and customized IoT hardware, AsiaRF is dedicated to pushing the boundaries of connectivity to empower businesses and industries globally.
SAN JOSE, Calif., Mar 10, 2026 — Synaptics Incorporated (Nasdaq: SYNA), today announced the SYN765x, an AI-native wireless solution that redefines Edge intelligence. As an industry-leading single-chip device combining AI-optimized compute with integrated Wi-Fi® 7, the SYN765x is designed to bring scalable, real-time intelligence directly to smart appliances, home automation systems, and Industrial IoT (IIoT) applications.
SYN765x integrates Wi-Fi 7, Bluetooth® LE 6.0, and Thread/Zigbee across 2.4, 5, and 6 GHz bands. Dedicated on-chip acceleration supports predefined AI-native control and signal-processing functions, reducing host processor load, while meeting strict latency and power requirements.
Built for high-performance wireless connectivity, the SYN765x pairs essential compute as a convenient, on-chip companion for smart devices at the Edge. Its single-chip integration significantly reduces system space requirements, simplifies design, and offers both engineering flexibility and cost savings.
By bringing advanced Wi-Fi 7 capabilities to low-power designs, SYN765x helps enable lower latency, seamless band switching, fast and secure reconnections, and access to the interference-free 6 GHz spectrum. Until now, power and cost constraints have limited adoption of the latest Wi-Fi standards in embedded systems. This solution meaningfully reduces these barriers, accelerating next-generation Wi-Fi adoption across Edge IoT devices.
The SYN765x also enables advanced wireless sensing, including presence detection, motion tracking, and proximity awareness using standard Wi-Fi and Bluetooth signals. For Wi-Fi sensing, it combines high-accuracy Channel State Information (CSI) extraction with on-device machine learning. This differentiated architecture is designed to deliver high accuracy and reliability, supporting a broader range of sensing use cases compared to other connectivity solutions in its class.
In addition, Bluetooth Channel Sounding enables accurate, power-efficient distance measurements under typical operating conditions — offering an alternative to more expensive technologies such as mmWave radar or ultra-wideband. Integrated logic, DSP, and NPU resources provide a high-performance platform for embedded Edge AI workloads, while AI-driven contextual awareness enhances power efficiency.
The SYN765x supports flexible deployment models. It can operate as a co-processor alongside a host application processor or MCU, or function in standalone and host-less configurations thanks to its generous on-chip memory and processing resources. By eliminating the need for a separate microcontroller in many designs, the SYN765x significantly reduces system complexity and cost—offering a new class of high-performance, battery-powered IoT devices.
“Intelligence at the Edge demands uncompromising wireless performance,” said Vikram Gupta, SVP & GM, Edge Compute & Connectivity Division, Synaptics. “SYN765x extends our leadership in wireless technology by integrating Wi-Fi® 7, BLE, Thread/Zigbee with AI-native processing. This simplifies system design, lowers power and cost barriers, and accelerates adoption of next-generation wireless across consumer, industrial, and enterprise IoT markets.”
SYN765x Features at a Glance:
Global customers and partners count on Synaptics to power the next generation of connected, AI-enabled IoT applications.
Gilles Drieu, Chief Technology Officer, ADT
“Connectivity is at the heart of everything we deliver to our customers. Our partnership with Synaptics enables us to build security solutions that are more reliable, intelligent, and responsive. Their portfolio of connectivity solutions gives us the flexibility and performance we need to design next-generation systems that protect homes more effectively, while remaining simple to deploy and manage.”
Janet Wei, CEO, Ampak
“Synaptics has been a trusted technology partner for many years, and our longstanding collaboration has been instrumental in advancing connectivity solutions across a wide range of applications. Together, we’ve consistently pushed the boundaries of performance, integration, and reliability, enabling customers worldwide to deploy robust, scalable wireless solutions and reinforcing our shared vision for the future of connected devices.”
Spencer Maid, President and CEO, Origin AI
“Synaptics brings world-class AI and connectivity capabilities that perfectly complement Origin’s AI Sensing platform. Together, we’re unlocking powerful new use cases that transform everyday environments into smarter, more adaptive spaces.”
Andrew Zignani, Senior Research Director, ABI Research, Strategic Technologies Team
“Artificial intelligence has been migrating out from data centers throughout the network, and now the Edge is truly ripe for local AI. From factory equipment, where reliability and security are key, to consumer electronics, where price sensitivity and privacy is acute, to wearables, where weight, size, and battery life are critical, Synaptics’ latest offering effectively balances and optimizes computing and reliable, future proof connectivity to better serve the evolving requirements and growing diversity of product types found at the network Edge.”
Mohit Agrawal, Global Practice Head for Edge AI and IoT, Counterpoint Research
“The shift from cloud-centric IoT architectures to real-time Edge intelligence is accelerating, particularly as latency, privacy, and bandwidth efficiency become critical design priorities. Integrating AI acceleration with Wi-Fi 7 on a single SoC represents an important step forward for the AIoT ecosystem. As Wi-Fi 7 adoption expands over the next several years, solutions that tightly couple on-device inference with high-performance connectivity will be well positioned to support next-generation smart home, industrial, and infrastructure applications.”
Availability
The Synaptics SYN765x solution is expected to begin sampling in the second calendar quarter of 2026, with production targeted at the last calendar quarter of 2026. Development kits are expected to be available for sampling in calendar Q2 2026.
In just a few years, AI has leaped from research labs to the real world. As this momentum accelerates, AI is no longer a vision, but a reality. One which is not confined to the cloud, instead quickly moving to the edge. Generative AI, agentic AI and physical AI are now being embedded directly into the devices and infrastructure that power our world. This imminent shift changes the game, especially for teams moving from prototype to production — success depends not on training the largest models, but on deploying AI that is efficient, secure and scalable where it matters the most. At Qualcomm Technologies, our edge AI processors portfolio is purpose-built for this future, designed to enable intelligence across industries, from early validation to scaled deployment.
That’s the difference between edge AI in theory and edge AI in production. Edge AI is where artificial intelligence matures from only running in centralized data centers to distributed, real-world deployments. Real-time decisions can’t wait for the cloud. Sensitive data can’t risk exposure. Billions of devices can’t rely on constant connectivity. Purpose-built edge AI processors solve all these challenges, embedding intelligence directly into the devices and AI appliances.
Our journey in industrial and embedded IoT began many years ago, culminating in the launch of our most advanced, industrial-grade Dragonwing IQ Series processors in 2024, including the Dragonwing IQ6, IQ8 and IQ9. These processors are designed to bring AI to the edge across factories, warehouses and infrastructure. In 2025, we expanded this portfolio with the Dragonwing IQ-X Series, powering the next generation industrial PCs running Microsoft Windows. Most recently at CES 2026, we unveiled our flagship Dragonwing IQ10 Series of industrial processors, marking our entry into advanced robotics.
Together, this comprehensive portfolio is designed to enable us to scale edge intelligence without forcing teams to re-architect as they grow across every industrial vertical and tier, supporting AI processing from 1 TOPS for sensor-level intelligence to 350 dense TOPS for complex vision and decision-making tasks. For customers requiring even greater AI capability, our AI on-premises appliances deliver peta-flop-class performance and can run 200B-parameter models on a single system. This enterprise-grade performance keeps data and processing local, which is an important capability for industrial applications with strict security, privacy, reliability and latency requirements.
At Embedded World 2026 in Nuremberg, we’re bringing more than 20 demonstrations powered by our industrial processors and AI on-prem appliances. These demos reflect the momentum we’ve built turning prototypes into deployable systems to drive digital transformation across industries, in close collaboration with our partner ecosystem, to make AI real and scalable. This theme is explored across distinct focus areas including developer enablement and solutions that span from commercial and industrial to robotics. If you are at the show, be sure to drop by our booth in Hall 5, Stand #5-161. If not, let me spotlight a selection of these areas to illustrate how edge AI is delivering value across multiple verticals and use cases.
Powered by the Dragonwing IQ8-Series processor, the Arduino VENTUNO Q combines Arduino’s developer-friendly ecosystem with industrial-grade edge AI performance of up to 40 dense TOPS. This new platform democratizes access to production-ready AI development, enabling engineers to rapidly build and deploy edge AI systems. To showcase its capabilities, we have multiple demos on display at the Qualcomm and Arduino booths, including an interactive smart mirror, an engaging chatbot, an AMR and more.
For the broader industrial use cases, we are bringing several demos in coordination with partners to highlight how connected intelligence can help enhance operational efficiency, safety and productivity. Below, let me spotlight four important use cases:
Collaborating with Qt Group and Edge Impulse, we are making great strides aimed at radically improving the development experience for AI-driven industrial software. Our demonstration exhibits distributed edge AI on multiple interconnected manufacturing sites. Powered by Dragonwing IQ9, the demo showcases four AI applications — visual defect detection, equipment fault detection, worker safety and AI factory assistant — with multiple AI models operating simultaneously. Additionally, we enabled Dragonwing IQ6-based Human Machine Interface (HMI) devices running on-device face detection for operator access and AI tethering with Dragonwing IQ9 to leverage AI factory assistant remotely. With AI tethering, operators using Dragonwing IQ6-based HMIs are not limited by the on-device AI and simple monitoring of the data, but they are able to leverage the power of Dragonwing IQ9 to help get prescriptive responses on the defects and manufacturing lines.
By analyzing visual data in real time, edge AI can be trained to detect defects, variations and anomalies early on the manufacturing line, improving production yield, reducing waste and helping to ensure consistent quality. In collaboration with CODESYS, we are demonstrating an industrial PC for PCB defect detection, running Windows on Dragonwing IQ-X that supports EtherCAT, motor control, image processing and AI chatbot. It also demonstrates Qt for user interface.
PLC and HMI consolidation
We demonstrate consolidation of PLC and HMI functions on a single Dragonwing IQ9 platform through virtualization. Here, Real-Time Linux is used to run CODESYS for PLC application enabling low-jitter motor control over EtherCAT, while virtualized Windows HMI on the same Dragonwing IQ9 platform shows the data for monitoring and control. This consolidation of various functions and devices over single platform through virtualization is designed to give customers flexibility to implement, scale, add and modify capabilities as the factory evolves. Additionally, consolidation can lead to reduced hardware and maintenance costs.
A core challenge in industrial networking is connecting a diverse set of sensors and devices that support different connectivity standards. Our Dragonwing IQ9 based industrial gateway is purpose-built to address this, supporting 5G, Wi-Fi, Bluetooth, LoRa, Ethernet and other communication standards, while designed to deliver powerful edge AI for sensor fusion and other edge AI workloads. Our demo aims to show how on-device vision processing can support worker safety, integrate with industrial SCADA systems to enable real-time production monitoring, and support simplified device management.
Edge AI redefines enterprise productivity and operations when it’s designed to deploy consistently at scale. With cutting-edge VLMs and LLMs, organizations can personalize interactions, improve efficiency and strengthen security, all while keeping data closer to where it’s generated. Here are a few demos that highlight edge AI in the commercial environment:
VLMs combine computer vision with natural‑language reasoning to unlock richer video insights, from scene understanding to intelligent action triggering. Running these models at the edge enables faster decisions and stronger data privacy. Our demo showcases multiple Dragonwing IQ9‑based solutions, including a video surveillance demo based on the Qualcomm Insight Platform, a performance comparison against a competitor’s solution and an example of multiple VLM instances running simultaneously on a single device.
Our Dragonwing processor platforms also enable a wide range of AI‑driven retail experiences designed to elevate customer engagement. In collaboration with BRICKS, we’re demonstrating an Dragonwing IQ9-based interactive digital signage solution that powers more intelligent, dynamic storefront interactions.
In collaboration with Consult Red, we’re also demonstrating a voice‑enabled fast‑food ordering kiosk that runs Android on Linux with the Dragonwing IQ9 industrial processor, featuring on‑device voice control powered by multiple AI models. Using Android on Linux, we can now run modern Android apps on a highly powerful edge AI industrial platforms that are also ruggedized and come with long-life support.
Announced at CES 2026, the Edge Impulse On-Prem AI can enable secure local AI inference and MLOps for a wide range of use cases. Customers can run inference for models up to 120B parameters, like GPT-OSS 120B, at just 150W on one Qualcomm Cloud AI 100 Ultra card. This architecture offers enterprise-grade precision, secure operation, private networks and full offline support.
The appliance also powers on‑premise generative AI enterprise workloads, as demonstrated through collaborations with context.ai for content creation. In telecom, we have supported Lanner and Ecrio to enable telecom and generative AI applications. We also collaborate with partners like Aramco for oil & gas and Siemens for industrial projects. To scale, we work with hardware vendors such as Aetina, Advantech, XSLAB and AUK Computing.
Our Dragonwing processors are helping to power the future of physical AI, where prototype-to-production gaps are most visible, to deliver end‑to‑end robotics solutions that scale intelligence from household robots to full‑size humanoids. Our demos showcase our work with partners and customers across a broad range of robotics use cases. We are also highlighting how we scale robotics development and deployment, from rapid prototyping on Dragonwing IQ8 with VENTUNO Q to an Advantech and Autocore AMR prototype, and more. Learn more about our latest robotics progress.
We are also launching Dragonwing Robotics Hub, built on the Arduino Project Hub, to accelerate this journey with ready‑to‑run samples and end‑to‑end workflows spanning sensors, Dragonwing compute, edge AI and robotics control, complete with schematics, documentation, tutorials and reusable sample code. Learn more about our latest robotics progress.
At Qualcomm Technologies, we’re not just talking about transforming industries, we’re leading it. Through our integrated portfolio of purpose-built edge AI solutions, we’re working closely with industry leaders to solve real-world challenges across industrial, commercial, robotics and beyond to help teams move from evaluation to deployment with confidence.
The momentum is accelerating and 2026 promises to be a pivotal year for scaling AI to the edge. Follow us for upcoming announcements as we continue pushing the boundaries of what’s possible.
In massive warehouses and factories around the world, fleets of autonomous guided vehicles (AGVs) and autonomous mobile robots (AMRs) are transforming the way goods move from shelf to shipment. Productivity can soar by 300%, but a single human step into the robot work zone can bring the entire operation to a halt. Every pause for safety costs time, money and momentum.
That’s the challenge Redpoint Positioning set out to solve—and they’re solving it with Qorvo Ultra-Wideband (UWB) technology.
“For large-scale automated warehouses, efficiency and safety have always been in tension,” explains John Calcio, Vice President of Partnerships at Redpoint Positioning. “If a technician needs to enter an AGV zone, regulations require shutting down every robot in the area. It’s safe—but it’s also massively inefficient.”
High-volume warehouse environments may deploy thousands of fast-moving robots working side-by-side with human crews. Maintaining over 99% reliability in such a dynamic, high-density space is critical, particularly when functional safety standards like SIL2 (Safety Integrity Level 2) must be met.
“The easiest way to stay compliant is to stop everything,” Calcio says. “But that’s a brute-force solution. We knew we could do better with real-time, centimeter-level awareness.”
Redpoint’s real-time location system (RTLS) integrates directly into AGVs and wearable safety gear. Each robot and worker is equipped with a UWB tag that continuously tracks its position with centimeter-level accuracy and millisecond latency. When a worker enters a restricted zone, nearby robots automatically slow or stop—without requiring a full fleet shutdown.
“Think of it as indoor GPS,” Calcio explains. “Our technology lets machines and people ‘see’ each other in real time, even in environments where Wi-Fi or vision systems struggle. It’s fast enough that a forklift can slow down or reroute before a collision is possible.”
The result is a SIL2-compliant wireless safety system that keeps both people and production moving.
At the heart of Redpoint’s RTLS is Qorvo’s UWB technology. With its low power consumption, precision ranging and robust performance in congested RF environments, Qorvo UWB provides the foundation Redpoint needed to meet industrial-scale performance and safety goals.
“Qorvo’s UWB delivers exactly what these environments demand—centimeter-level accuracy, millisecond response and exceptional reliability in complex, high-density deployments,” says Percy Yu, Qorvo Regional Marketing Manager in China. “Redpoint has been a long-term partner of Qorvo—over seven years now—and they’ve built on our UWB platform to create a proven, safety-certified RTLS solution.”
That partnership has already paid dividends: Redpoint’s system now powers tens of thousands of AGVs worldwide, improving operational efficiency by an average of 4–5%, with up to 15% increases in operator productivity and a 95% reduction in workplace accidents.
“These numbers may sound modest,” Calcio notes, “but when you’re talking about multi-million-square-foot facilities, a few percentage points translate to tens of millions of dollars in savings.”
One of Redpoint’s key differentiators is scalability. Their RTLS solution is purpose-built for environments where hundreds or thousands of moving assets share the same wireless space.
“When you have that many nodes transmitting position data at once, congestion and interference can become real issues,” says Calcio. “We engineered custom wireless protocols to keep latency and reliability at SIL2 standards, even in massive deployments.”
And that scalability isn’t limited to warehouses. Redpoint’s technology is now being adopted in industries ranging from logistics to mining—and they’re not alone in this UWB-powered expansion.
In China, Tsingoal—the country’s largest RTLS solution provider—is leveraging Qorvo UWB across a diverse portfolio that includes robotics control, factory automation, mining, logistics, parking systems and smart city infrastructure. Tsingoal’s UWB-based systems provide real-time tracking of assets, vehicles and personnel, improving safety, workflow efficiency and spatial awareness across industrial and urban environments.
From intelligent manufacturing lines to airport logistics and even humanoid robotics demonstrations, Tsingoal’s implementations showcase the versatility of Qorvo’s UWB technology in complex, multi-layered operations.
Both Redpoint and Tsingoal exemplify how Qorvo’s UWB technology is enabling a new generation of connected, intelligent and safer industrial ecosystems.
“What’s exciting,” says Yu, “is that we’re seeing global adoption. From North America to Asia, companies are realizing that UWB isn’t just about precision—it’s about trust, reliability and safety at scale.”
Calcio agrees. “This is where the market is going. The future of automation depends on real-time awareness—and that starts with UWB.”
Find out more about Qorvo UWB solutions here.
It often starts quietly.
A customer-facing AI assistant hesitates before responding.
An automated workflow pauses, then resumes.
A recommendation engine delivers inconsistent results—right one time, wrong the next.
Nothing is technically “down.”
No alerts are firing.
But confidence begins to slip.
Teams look first at the model. Then the data pipeline. Then cloud capacity. Everything appears healthy—until someone asks the uncomfortable question:
Could this be the network?
Across large, globally distributed enterprise networks, this pattern is emerging with increasing consistency. As organizations embed AI into core business workflows—customer engagement, software development, security operations, supply chain optimization—the network is being asked to support workloads it was never originally designed for.
Clearly understanding the limitations of your existing architecture can help you anticipate challenges before they impact operations, refine deployment strategies, and establish safeguards that prevent costly disruptions. This will enable smoother AI adoption and drive more reliable and successful technology outcomes for your organization. So, let’s examine AI workloads and where conventional networks struggle.
One of the most common missteps enterprises make is treating AI workloads like traditional applications.
They’re not.
AI workloads are highly sensitive to latency, intolerant of jitter, and dependent on continuous, real-time data movement across campuses, branches, clouds, and edges. They introduce new traffic patterns—east-west, north-south, machine-to-machine, agent-to-agent—that many existing network designs were never optimized to observe or assure.
In an AI-driven workflow:
When performance degrades—even slightly—the impact isn’t just slower response times. It shows up as inconsistent outcomes, unreliable automation, and hesitation to trust AI-driven decisions.
Networks built for predictable applications don’t fail catastrophically here.
They struggle inconsistently—which is harder to diagnose and more damaging at scale.
Traditional network performance models assume:
AI breaks all three.
Traffic shifts dynamically based on where inference occurs. Application behavior changes in real time. Congestion doesn’t appear as a clean outage—it surfaces as erratic AI behavior that’s difficult to reproduce or explain.
Operations teams are left asking:
Many existing monitoring tools struggle here, but they report utilization, not experience. Health, not intent. Metrics without the context needed to explain why AI outcomes fluctuate.
The lack of insight is inevitably paired with the following result:
AI workloads run—but rarely deliver consistent performance as they scale.
Before AI, network teams relied on assurance to gain end-to-end visibility and pinpoint network issues impacting user experience.
In an AI-driven world, assurance becomes foundational, providing dynamic, continuous monitoring and proactive management to keep pace with the complexity and speed of AI workloads.
AI systems depend on continuous confidence that:
Networks designed for manual intervention rely heavily on after-the-fact investigation. Humans piece together logs, dashboards, and alerts across multiple tools and teams.
That approach doesn’t hold when AI systems operate continuously and autonomously.
AI doesn’t wait for tickets.
AI doesn’t pause for triage.
When visibility and trust degrade, AI systems don’t stop—they make poorer decisions.
Without assurance integrated into the network itself, organizations often slow AI adoption—not because the use cases lack value, but because outcomes become unpredictable.
Security was historically designed to protect human-driven applications moving at human speed.
AI operates at machine speed—and it exposes every point of friction in between.
Many traditional security approaches rely on:
That friction was manageable for human-driven applications. For AI workloads operating continuously and autonomously, it becomes a limiting factor.
Every additional hop adds latency.
Every policy mismatch introduces unpredictability.
Every blind spot increases risk.
When security isn’t integrated directly into the network fabric, teams are forced into trade-offs they shouldn’t have to make—between protecting the environment and keeping AI responsive.
Performance, assurance, and security challenges are symptoms. The underlying constraint is architectural.
Most enterprise networks evolved as collections of domains:
Each optimized independently. Each managed with its own tools, policies, and operational workflows.
AI workflows span all of them—simultaneously.
They require shared context, coordinated policy enforcement, and the ability to reason across domains in real time. When architecture remains fragmented:
This is why many AI initiatives stall after early success. The models work. The pilots prove value. But scaling exposes friction—not in AI itself, but in the network layers beneath it.
As AI moves from experimentation to everyday operations, a pattern is becoming clear.
AI doesn’t struggle because models lack sophistication. It struggles because the networks they run on were designed for a different operating model.
Networks optimized for predictable, human-driven applications need to support continuous, autonomous, and outcome-driven workflows.
For many organizations, this realization doesn’t arrive as a dramatic failure. It surfaces through inconsistency, operational friction, or difficulty scaling what initially worked. Over time, these signals accumulate—prompting a broader rethinking of how the network fits into the AI roadmap.
Your AI roadmap can’t wait for pressure to build. In the years ahead, as AI becomes embedded into every workflow and decision loop, networks will increasingly be judged not just on availability, but on their ability to assure outcomes at machine speed. The time for recognition and action is now.
Because in the AI era, the network isn’t just infrastructure.
It’s part of how intelligence moves, reasons, and delivers value.
WASHINGTON, D.C. March 17, 2026: Today DNSFilter announced the launch of CyberSight, a new behavioral analytics and intelligence capability within the DNSFilter platform designed to move beyond DNS filtering and deliver full visibility into user activity across devices, applications, and web traffic.
With organizations now using an average of more than 100 SaaS applications, security teams lack visibility into who is using what applications and what threats are impacting their networks. Security teams and managed service providers (MSPs) face a growing challenge: traditional DNS filtering only reveals what was blocked, leaving critical blind spots around what users actually do online. Without visibility into real user behavior, organizations struggle to investigate incidents, identify risky SaaS usage, and detect emerging threats.
“DNS filtering has always been excellent at stopping threats, but security teams also need visibility into the activity that never gets blocked,” said Ken Carnesi, Chief Executive Officer at DNSFilter. “CyberSight delivers that missing context, giving organizations the ability to understand user behavior across the web, uncover shadow IT, and respond to incidents significantly faster. It’s a powerful new layer of intelligence within the DNSFilter platform.”
Architecturally integrated within the platform’s Pro and Enterprise tiers, CyberSight is a purpose-built user behavior analytics solution. By delivering clear, digestible insights into full user activity including what happened, when it occurred, and how long interactions lasted CyberSight enables organizations to move beyond reactive blocking and proactively reduce SaaS risk.
“CyberSight provides us with a much-needed lens into user data and behavioral patterns throughout the workday,” said Thomas Connolly, IT Manager at Crescent Crown. “By pairing these insights with DNSFilter’s Protective DNS, we’ve created a layered defense that makes it easy to spot and mitigate risky behaviors. This combination has fundamentally shifted our security posture, allowing us to understand the context behind user actions and respond significantly faster.”
Unlike traditional DNS filtering solutions that only log blocked requests, CyberSight captures activity across clicks, applications and full URLs, including IP addresses typed directly into browsers. The solution separates meaningful user actions from background system noise, providing security teams with clear behavioral timelines and actionable insights.
With CyberSight, organizations can:
Accelerate incident response
Detailed event timelines and granular logs help security teams quickly investigate suspicious activity and understand user behavior leading up to an incident.
Detect shadow IT and risky SaaS usage
CyberSight reveals previously hidden application usage, helping organizations identify unauthorized services and enforce security policies.
Optimize SaaS spending
Visibility into application activity helps IT teams identify redundant tools, eliminate unused licenses and better manage software investments.
Early adopters are already seeing the value of this additional visibility. CyberSight is currently available for Windows devices, with macOS support planned for the second half of 2026. Learn more about DNSFilter’s CyberSight offering here.
DNSFilter will be exhibiting at this year’s RSA Conference in San Francisco. Visit https://explore.dnsfilter.com/rsac to book a meeting.
In the embedded systems world, embedUR has spent the last twenty years in the trenches. We’ve built the wireless infrastructure that runs enterprise access points and the residential gateways sitting in millions of homes. In that time, we’ve learned one thing: regulatory shifts aren’t just legal problems. They are engineering problems.
The EU Cyber Resilience Act (CRA) is one of the biggest shifts we’ve seen in a decade. It isn’t just a footnote in a policy briefing; it’s a fundamental change in how we need to or should design, build, and most importantly, sustain products. At embedUR, we’ve guided teams through every Wi-Fi evolution from the early days of 802.11n to the complexities of Wi-Fi 6E and the new world of tinyML at the edge.
From where we stand, the CRA shouldn’t be a headache. It should be the catalyst that finally forces the industry to build smarter, more resilient systems.
But let’s be honest: for an engineering lead trying to hit a product launch date while staring down resource gaps, the CRA looks like a mountain of paperwork. To climb it, you have to bridge the gap between using an open-source RTOS like Zephyr and meeting the rigorous demands of European law.
We are currently in early 2026. The grace period is evaporating. The CRA was formally adopted back in late 2024, and the first major hammer will drop on September 11, 2026. That is the date when mandatory vulnerability reporting goes live.
If your device has a digital element and you sell it in the EU, you have 24 hours to report an exploited vulnerability. Not 24 hours after you fix it—24 hours after you find it.
By December 2027, the remaining requirements kick in: security-by-design as a legal mandate, machine-readable Software Bills of Materials (SBOMs), and a 5-year support window. If you miss these marks, the fines are eye-watering: up to €15 million or 2.5% of your global turnover. For any U.S. company exporting to Europe, which is a massive chunk of the global market, this is now the cost of doing business.
We’ve been working with Zephyr RTOS since its early releases in 2016. It’s under the Linux Foundation, it’s transparent, and it’s modular. On paper, it’s the perfect answer to the CRA.
Zephyr’s PSIRT (Product Security Incident Response Team) is excellent. They handle CVEs (Common Vulnerabilities and Exposures) with a level of transparency that mirrors the Linux kernel. When a bug is found in the networking stack, the fix is public and documented.
Furthermore, the Long-Term Support (LTS) releases, like the 3.7 LTS, offer five years of security maintenance. That perfectly maps to the CRA’s support requirements.
Zephyr also gives us the “West” tool. Using west spdx, we can generate an SBOM automatically during the build process. If you’ve ever tried to manually track every library and sub-component in a proprietary RTOS, you know that automated SBOM generation is a godsend. It takes the drudgery out of compliance.
However, this is the local engineer’s truth: upstream Zephyr is not the same thing as a shipping product.
Here is where the theory hits the floor. In our work sustaining Wi-Fi products for major global brands, where we manage fleets of over 10 million units, we’ve seen divergence as a major problem.
An engineer starts with a clean Zephyr LTS branch. Then, the silicon vendor provides a HAL (Hardware Abstraction Layer) that isn’t quite right for the specific power constraints of a Cortex-M4 gateway. So, the engineer tweaks the kernel. Then the marketing team requests a specific feature that requires a custom driver. Before you know it, you’ve forked the repo.
When a security patch comes down from the Zephyr upstream six months later, it doesn’t fit. It breaks your custom power management. It conflicts with your Wi-Fi driver. Now, compliance means a three-week rebase and a full regression test.
If you have a multi-SKU product line, maybe some devices use ARM and others use RISC-V, your vulnerability tracking turns into a nightmare. There are cases where a field-reported bug traced back to an unpatched erratum in a silicon vendor’s HAL. Because the team had drifted too far from the clean upstream code, they couldn’t pull the fix.
This kind of delay is exactly what the CRA will penalize.
The complexity doubles when you add tinyML. We’re seeing more industrial gear using TensorFlow Lite Micro on Zephyr to do predictive maintenance at the edge. These models are ultra-constrained. Every byte of RAM matters.
When you’re fighting for every kilobyte to make an AI model work on an IoT sensor, security overhead feels like an enemy. But under the CRA, that edge intelligence is part of the product’s digital footprint.
If your ML inference engine has a buffer overflow vulnerability, you are liable. Shoehorning these models into Zephyr while maintaining a secure, updatable bootloader (like MCUboot) requires a level of memory management that most teams aren’t prepared for.
So, how do we stop compliance from being a burden and turn it into a reason why customers choose you over a competitor? You have to treat security as a feature, not a chore.
The biggest mistake we see is that engineers treat Zephyr like a one-off download. You have to stay aligned with the LTS branches. We recommend a “cherry-pick” strategy. You pin your project to a stable LTS version and build a rigorous process for pulling in only the security patches you need. This keeps your custom “secret sauce” separate from the core RTOS plumbing.
If an auditor asks for your SBOM, you shouldn’t have to spend a week in Excel. We help our clients integrate SBOM tools directly into their CI/CD pipelines (like GitLab or GitHub Actions).
Every time a developer pushes code, the system should automatically run a vulnerability scan (using tools like grype or trivy) and generate a new SPDX file. If a vulnerability is found, the build should fail before the code ever reaches a device.
To handle the multi-SKU headache, we use Zephyr Device Tree overlays. Instead of having different codebases for different hardware, you keep one core application and use overlays to describe the hardware differences. This minimizes code drift and ensures that a security patch for your BLE stack works across every device in your catalog.
You can’t trust a simulation for CRA compliance. You need automated test rigs where real hardware is subjected to simulated exploits. At embedUR, we’ve built environments where we test everything from LTE stack vulnerabilities to timing glitches in crypto libraries.
This is how you catch a DoS (Denial of Service) bug before it gets found in the field and becomes a mandatory 24-hour report to the EU.
This is what we do. We aren’t a compliance consultancy that just hands you a PowerPoint. We are an engineering firm with over 400 expert embedded systems engineers who live in the kernel. Our architects have upstreamed code to Zephyr’s networking modules. We also know the silicon vendors—NXP, Infineon, Nordic, and Silicon Labs—and where their HALs usually break.
We have the sustainment framework ready to go. We handle the kernel upgrades, the CVE monitoring, and the SBOM maintenance so that your senior engineers can focus on the stuff that actually makes you money, like your next-gen predictive maintenance algorithms or your cloud-to-edge analytics.
The September 2026 deadline is closer than it looks on the calendar. In the world of embedded systems, eighteen months is one design cycle. If you start thinking about the CRA during the testing phase, you’ve already lost.
Is your Zephyr setup a solid foundation, or is it a house of cards? If you’re building for the EU market and worried about staying compliant without killing your product roadmap, let’s talk.
We can do a technical review of your current stack and show you exactly where the gaps are. We’ve been through the Wi-Fi wars and the IoT explosion; we can get you through this, too.
Let’s get your engineering team back to innovating. Schedule a Technical Review: Talk directly with our lead architects about your current Zephyr stack and CRA gaps.
