Axelera AI

For a long time, the hard part of an edge AI project wasn't the idea. It was the plumbing. Building the pipeline, the YAML, the model wiring, and waiting for hours before anything ran on hardware. The all-new Voyager Wingman takes that part off your plate. You describe what you want to build, it generates a working pipeline for Metis, runs it on the silicon, and iterates until it works. Which leaves the more interesting question: what would you actually build, now the pipeline can take care of itself? That's the idea behind The Prompt, Axelera's latest community project challenge. Pitch us a project, and the ten most creative ideas each get a powerhouse Dell Technologies Pro Slim Plus XE5 with a 16GB Metis PCIe card to keep, free access to Voyager Wingman, and one month to build. Winner takes €1,000, runner-up €500. Already have Axelera hardware? There's a community category too. Build with Wingman and the best entry wins an Axelera AI Mini PC. You don't need to have built anything to enter. You just need an idea worth building. Proposals are already landing, and they're seriously creative. Pitches close soon, so share your idea right now: https://lnkd.in/gqpU7M-w #EdgeAI #ComputerVision

Porting existing AI models to new hardware architectures often involves manually re-writing configurations, re-compiling models, and adjusting pipelines by hand. With Voyager Wingman, transferring computer vision projects from platforms like Nvidia and Hailo over to Axelera AI hardware is designed to be a 1-click process. Wingman translates and configures existing setup files automatically, saving significant integration time. Here is what else is covered in the video walkthrough: - Creating for Non-Engineers: How developers, domain experts, and non-engineers can build functional computer vision pipelines using plain-language prompts. - Debugging & Diagnostics: Fast troubleshooting for device enumeration, configuration errors, and compiler issues. - Execution & Summary: Running pipelines live on Metis hardware and key takeaways for streamlining edge AI development. Find the link in the comments to watch the video to see the workflow in action. #AIDevelopment #EdgeAI #ComputerVision #AIAccelerator #EmbeddedSystems #DeveloperTools #AgenticDev

GenAI Works put together a great list of 30 AI tools you need to watch in 2026. The thread running through all of them is the gap between AI that THINKS and AI that PERFORMS. Models reason, write, and plan better every month. Getting them to act in the real world, in real time, next to the data, is the harder part. Every tool on that list closes some aspect of that gap. Axelera sits at the far end of it, and we're kind of the odd one out. Almost everything on the list is software, while we're the silicon underneath. Metis runs vision and language models directly on the device, in the camera, on the machine, so AI can act on what it sees rather than just describe it. Foundational tech rather than the superstructure AI that's built on top of it. #EdgeAI #ComputerVision

Yesterday's World Cup final between Spain and Argentina showed exactly why real-time sports analytics require immense computer vision performance. With Spain dominating possession and launching twenty shots on goal, a standard broadcast console has to process massive amounts of concurrent spatial data just to track player positioning, defensive heatmaps, and tactical transitions. Handling a multi-stream workload this intense usually takes weeks of manual hardware optimization and video pipeline design. We wanted to simplify that process. Our latest walkthrough demonstrates how to build and optimize a soccer analytics layout using Voyager Wingman. Instead of writing custom configuration code from scratch, a developer can type a natural language prompt to automatically generate a complete development workflow tailored to the Voyager SDK and our edge hardware. Whether your goal is tracking tactical spacing for active broadcasts or pulling performance metrics from historic tournament footage, this workflow lets you move from a conceptual design to physical deployment at the edge in minutes. The full walkthrough video and supporting documentation are live on our developer platform. Watch the video: https://lnkd.in/drrs22Dj Join the discussion and try Voyager Wingman: https://lnkd.in/dWfm78Q6 #WorldCup #SoccerAnalytics #ComputerVision #AIHardware

Ten builders. A Dell workstation with a 16GB Metis PCIe card each, yours to keep. One month to build. 🚀 The Prompt is our new community challenge, built around Voyager Wingman. Describe the AI project you want to make, and Wingman generates the pipeline, runs it on Metis, and iterates until it works. You bring the idea, it handles the wiring. Pitch your project today, and we'll pick the ten most creative and ship each one a Dell Technologies Pro Slim Plus XE5 with Axelera inside. Winning build takes €1,000, runner-up €500. Already have Axelera hardware? The community category is open to everyone, and the best build wins an Axelera AI Mini PC. Pitches close soon, so get your idea in before the window shuts. Full details right here: https://lnkd.in/eQ8GSWM7 #EdgeAI #AIHardware

The Prompt hackathon is now live. Miss it and you're watching everyone else build amazing things. Visit Axelera now to pitch your AI project idea, and if it's one of the ten most creative, we ship you a powerhouse Dell Technologies Pro Slim Plus XE5 with a 16GB Metis PCIe card to build it on. You build with Voyager Wingman, so all you need to do is describe what you want and it generates the Metis pipeline, runs it on the hardware, and iterates until it works. Ideas from you, YAML and coding legwork by Wingman. One month to build. There's €1,000 for the best project, €500 for the runner-up, and if you already own Axelera hardware you can join the community category even if you aren't selected, with an Axelera AI Mini PC for the best entry. Pitches close 29 July. Final builds due 2 September. Go be a maverick! Link in the comments 👇 #EdgeAI #ComputerVision

Pitch us an AI project idea and you could win a Dell Technologies Pro Slim Plus XE5 with a 16GB Metis PCIe card to build it on. That's The Prompt, our new AI project challenge ✈️ You even build it with the all-new Voyager Wingman. Describe what you want, it generates a Metis pipeline, runs it on the hardware, and keeps iterating until it works. You bring the idea; it handles all the plumbing and pipe work. We'll pick the ten most creative pitches and ship each builder a Dell XE5 to keep. One month to build. €1,000 for the best project, €500 for the runner-up. Already got Axelera hardware? You can build something with Wingman, post it by the deadline, and the best entry wins an Axelera AI Mini PC. Pitches are open now, but close 29 July, so hurry up. Go be a maverick! Link in the comments 👇 #EdgeAI #ComputerVision

We're #hiring a new New Product Introduction Engineer in Greater Bristol Area, United Kingdom. Apply today or share this post with your network.

If you work in AI, build with AI, or just can’t stop thinking about it, HumanX is where you want to be! I’ll be on stage talking about decentralized AI: how it unlocks sharper insights while cutting energy costs dramatically, for organizations and individuals alike! Sharing the stage with some of the sharpest people in the field is a real privilege! Attending? Come find me! I’d love to share ideas on where AI infrastructure goes next. See you there! #AxeleraAI #DecentralizedAI #HumanX #SustainableAI