Recent Topics

Buongiorno a tutti, Mi chiamo Vito e sono il titolare di una società Italiana, una start-up. Insegno anche in un istituto tecnico di Brescia Robotica, Elettronica e Informatica. Abbiamo avviato un progetto per la realizzazione di un robot umanoide nella mia società con alcuni dei miei studenti, qualche mese fa nell'ambito dell'assistenza. Ho visto l'intervista dell'Ing. Del Maffeo e sono rimasto colpito. Spanner, ​@Spanner dobbiamo implementare il robot con AI e quindi di una scheda potente, possiamo collaborare? Il potenziale è grande..  come potremmo approfondire l'argomento? Vi ringrazio a tutti in anticipo.. 

Hi,I'm having trouble installing the metis-dkms driver. Can you help me?The device is a NanoPC-T6 from Friendly Elec, and the operating system is Ubuntu 22.04, also provided by Friendly Elec.I'm leaving the error message and OS information below. If there is any more information I should provide, please let me know. Thank you. The Error:$ sudo dpkg -i metis-dkms_0.07.16_all.deb(Reading database ... 155441 files and directories currently installed.)Preparing to unpack metis-dkms_0.07.16_all.deb ...Deleting module metis-0.07.16 completely from the DKMS tree.Unpacking metis-dkms (0.07.16) over (0.07.16) ...Setting up metis-dkms (0.07.16) ...Loading new metis-0.07.16 DKMS files...Building for 6.1.99Building for architecture aarch64Building initial module for 6.1.99ERROR (dkms apport): kernel package linux-headers-6.1.99 is not supportedError! Bad return status for module build on kernel: 6.1.99 (aarch64)Consult /var/lib/dkms/metis/0.07.16/build/make.log for more information.dpkg: error pro

Hello,I’m running inference with model  unet_fcn_512-cityscapes with pipe torch-aipu on the aetina eval board. It runs at 1.8 fps system and with 500ms of latency although device fps shows  11.5 fps capability. In the doc it is also mentioned that it should reach 18fps. I originally thought it was an issue with time wasted to load and decode png images from the SD card so I put them in shared memory but result are identical.I also tested yolov5s-v7-coco that should reach 805 fps but I can only achieve 214fps. Here are the output of: AXELERA_USE_CL_DOUBLE_BUFFER=0 ./inference.py yolov5s-v7-coco media/traffic3_720p.mp4 --show-stats --no-displayINFO : Deploying model yolov5s-v7-coco for 4 cores. This may take a while...|████████████████████████████████████████| 12:41.1 arm_release_ver: g13p0-01eac0, rk_so_ver: 9========================================================================

Problem SummaryPlatform: ARM64 (RK3588-based boards like Radxa ROCK 5B Plus)OS: Ubuntu 22.04Issue: GStreamer plugins (axinplace, axtransform, etc.) fail to build despite:All GStreamer development packages installed CMake detecting GStreamer successfully (GST_FOUND=1) No explicit build errors or warnings Source code present in operators/gstaxstreamer/Symptom: ERRORfailed to create element of type axinplace (decodebin-link0)when trying to do inferencebashgst-inspect-1.0 axinplace# Returns: "No such element or plugin 'axinplace'"After build, libgstaxstreamer.so is missing from operators/lib/ even though the build claims success.Root Cause AnalysisThe Silent Dependency FailureThe issue is a silent CMake conditional failure caused by a missing platform-specific dependency that's not documented anywhere in the Voyager SDK.What the developers missed:Incomplete dependency documentation - The SDK docs list standard GStreamer packages but omit hardware-specific requirements Silent build failure

Environment SoC/Board: Rockchip RK3588, Radxa ROCK 5B Plus    . It has 2 M.2 slots! OS/Kernel: Debian bookworm, 6.1.43-15-rk2312 (aarch64) RAM: 16GB (also tested with mem=4G kernel parameter; total reported ~3.8Gi) Device: Axelera Metis PCIe accelerator, PCI ID 1f9d:1100 (shows as “Axelera AI Metis AIPU”) PCIe topology: Endpoint at 0001:11:00.0 behind upstream bridge 0001:10:00.0 (Gen3 x2) Drivers: axl/metis-dkms 1.2.3; “Kernel driver in use: axl” shown; /dev/metis-1:11:0 device node present when driver bound Container: Voyager SDK tested in Ubuntu 22.04 container (privileged); device access tests moved back to host after hangs observed inside container Device-tree overlay (added to expand ranges window to 48MB)Reason: Default 32-bit non-prefetchable MEM window under fe160000.pcie was ~14MB; Metis BAR2 asked for 32MB (plus headroom), so a 48MB window was used in different place that seems free. Overlay below is applied via extlinux overlays:text/dts-v1/;/plugin/; / { compat

Hi all,I am trying to compile a high-resolution model (1504×1504) on a Metis PCIe card, but the compilation fails due to a memory constraint:ERROR : RuntimeError: Could not find a tiling factor that fits the memory constraints l1_constraint=4011520 l2_constraint=7894528. After attempt=7 and h_size=1 and adj_factor=1, memory usage still is memory_usage={L1: {190: 141376, 193: 5017600, 191: 143360}} and per-pool memory usage {L1: 5302336}.The error indicates a required L1 size (≈5.3 MiB) exceeds the constraint (≈4.0 MiB).My questions are: Voyager SDK Solution: Is there a Voyager SDK flag or setting to adjust the tiling to overcome this ≈4 MiB L1 limit? Hardware Solution: Are there any Metis card variants (or multi-AIPU cards) that feature a larger per-core L1 cache? Any guidance on compiling this high-resolution model is appreciated. Thank you!

Hello guys;I did not follow the instructions in the link below on the Orangepi 5 device, but it does not see the device.“ERR DEVCICE NOT FOUND”help! https://support.axelera.ai/hc/en-us/articles/27059519168146-Bring-up-Voyager-SDK-in-Orange-Pi-5-Plus

Hello everyone, I’m trying to install some packages using APT with my Axelera Metis Compute Board with Voyager Linux 1. However the /etc/apt/sources.list is empty. What kind of repositories should I use?Are there any tutorials?Thanks.Best regards

Hello,I’m trying to test the AxInferenceNet C++ Integration.When I launch the test with the Voyager SDK’s given code and the yolov8s-coco model, the results of the detection are totally staggered. Display of the result :I followed the AxInferenceNet C++ Integration Tutorial and I’m using Voyager SDK with Ubuntu 22.04.Could someone help me to understand the results and give me a bunch of idea to solve it ? Thanks in advance.

Hello,I created a request from support.axelera.ai and I’m part of a team, how can i gave access to my request to my team?I saw that there is a tab in request “Request I’m CC’d on” so i think there is some implementation for this but i can’t find it anywhere.

Dear team, Is it possible to use hardware acceleration provided by the chip for encoding into x264 or x265 video stream on gstreamer. I was able to find plugins here but no specific mention for hardware acceleration.Thank you.

I get this error when I try to deploy Real-ESRGAN. Can anyone tell me why?./deploy.py /home/aetina/Desktop/voyager-sdk-release-v1.4/ax_models/zoo/torch/real-esrgan-x4plus-onnx.yaml --num-cal-images 50INFO    : Detected Metis type as m2INFO    : Compiling network real-esrgan-x4plus-onnx /home/aetina/Desktop/voyager-sdk-release-v1.4/ax_models/zoo/torch/real-esrgan-x4plus-onnx.yamlINFO    : Compile model: real-esrgan-x4plus-onnxINFO    : Using representative images from /home/aetina/.cache/axelera/data/coco2017_repr400 with backend ImageReader.PIL,pipeline input color format ColorFormat.RGBINFO    : Prequantizing real-esrgan-x4plus-onnx: real-esrgan-x4plus-onnxINFO    : Quantizing network real-esrgan-x4plus-onnx /home/aetina/Desktop/voyager-sdk-release-v1.4/ax_models/zoo/torch/real-esrgan-x4plus-onnx.yaml real-esrgan-x4plus-onnxINFO    : Compile model: real-esrgan-x4plus-onnxINFO    : Using representative images from /home/aetina/.cache/axelera/data/coco2017_repr400 with backend ImageRead

Final Submission Notes - “Elderly Guardian” (edge fall-detection prototype)I had a blast building this. Between my daughter’s sports games this weekend, I squeezed in some late-night tinkering to land a working demo I’m proud of. The YouTube link in my submission shows the prototype end-to-end: live pose detection, on-screen overlays, and a smart-plug alert when a real fall is confirmed.What I builtA privacy-first, on-device fall detector that runs YOLOv8 Pose on an RK3588 with a Metis (Voyager SDK) accelerator, watches a Logitech C920 feed, and triggers a TP-Link Kasa plug to kick off downstream alerts/automations to an Amazon Echo DotHardware + stack Compute: RK3588 SBC Accelerator: Axelera Metis M.2 (via Voyager SDK) Camera: Logitech C920 (tested at 1280×720 @ 30 FPS) Network: Edimax EW-7822UAC USB Wi-Fi Alert path: TP-Link Kasa smart plug (local control with python-kasa) Model: YOLOv8 Pose (Ultralytics weights) running through Voyager’s pipeline How it works (quick tour)

Hi there,i´m working with an Axelera M.2 chip at uni. For some research on green AI i´m doing i want to track the power usage of the chip. As far as i know i need to track the power through the PCIe connection. Is there a way to track it with the AIPU directly or do i need to use the PCIe? And if there is a way can i save the usage in an array easily or is this a rather tricky situation?Thanks in advance Morruebe

Introduction Setting up the Linux System Running a First Demo Obtaining a Dataset Adding Custom Images to a Dataset Fine-Training the YOLO Model using the Dataset Exporting the Model Configuring the Pipeline Compiling the Models Running the Pipeline Adding a Third Stage Connecting to the Real World Tips / Tricks Summary IntroductionThe Automatic Number Plate Recognition (ANPR-for-All) project is intended to simply provide an easy-to-use vehicle license plate detection system for building into all manner of solutions. For instance, you may wish to register certain license plates for automatically opening gates, such as for emergency services or for security, or for displaying alerts on an LCD screen when certain vehicle license plates are detected.Check out the video here: To get there, I had a rather nice Axelera Metis M.2 Artificial Intelligence Processing Unit (AIPU) in an M.2 form-factor (you can see it toward the top of the photo above), installed on a Linux single-board computer (

Hello, I’m new to voyager SDK and everything related to AXELERA and i need help deploying a custom onnx model that has 2 inputs and 20+ outputs, all i want for now is to manage to run an inference with my own dataset/image and to save to disk the raw outputs of the model. The input is a pair of 2 files. I have data in yuv format and it’s saved in input_y.npy and input_uv.npy. Do you have any idea how can i do this or if it’s possible from the voyager 1.3 or do i need to add a split layer so i need just 1 input and only then i can deploy it?  Or to upgrade the voyager version to 1.4 

Will Metis M.2 ( AXE-BME20M1AR01B02) work with QNAP NAS?

Does anyone have any suggestions on how to implement SAHI inference with Yolo via voyager-sdk to run on metis m2?

Our development team requires specialized expertise for Axelera AI platform implementation and edge AI acceleration optimization. We're seeking an experienced consultant with 5-7weeks availability to support us with

Thought you might want to check out the live stream from YOLO Vision that’s happening today in London. Some of the Axelera AI team are there, and will be up on stage soon! 

Hello,I am currently running performance tests on the Metis M.2 card with different models (e.g., YOLOv8n).I observed a behavior that I would like to clarify: When running with 1 AIPU core, I get lower latency and the same FPS as when using multiple cores (4). However, with 4 AIPU cores, the latency per image is higher, even though the FPS does not increase. In addition, CPU utilization is noticeably higher with 4 cores (around 17%) compared to only ~0.8% with 1 core. I would like to know: Is this expected behavior (due to synchronization overhead between AIPU cores)? Is there a recommended configuration to optimize either latency or throughput (FPS) depending on the use case? Is it normal that CPU usage increases significantly when using 4 cores, given that processing is supposed to be largely handled by the AIPU? Thank you in advance for your clarifications.Best regards,

Hello everyone,I have been trying to compile the PRL Tracker model (see link below) for about a month using both Hailo and Vitis AI, but so far I have not been able to fully meet the requirements.I am now considering using the Metis Accelerator, but I am not sure whether I would run into the same issues.Has anyone here successfully deployed this model on the Metis Accelerator, or could you confirm if it is supported?Repository: https://github.com/vision4robotics/PRL-Track/tree/main Any guidance or shared experience would be greatly appreciated.Thanks in advance! 

PART ONEIntroduction & ForewordRecent advancements in edge AI hardware and software have opened up exciting possibilities for intelligent, compact systems that were unimaginable just a few years ago. Among these innovations, Axelera’s Metis devices stand out—delivering exceptional AI throughput in a small form factor and low power envelope, enabling use cases that were previously out of reach.For the past few years, I’ve had the privilege of living by the sea, with a panoramic view of the vibrant marine activity unfolding daily. From cargo ships to fishing boats, leisure yachts to water scooters, the constant movement on the water has always fascinated me. As someone passionate about extracting insights from data, I’ve often wondered about the patterns and stories hidden in this daily marine traffic. Yet, as a human observer, I’m limited in how much I can monitor or record—leaving my curiosity largely unsatisfied.Beyond curiosity, I’m deeply committed to the preservation of our mar