Skip to main content

Recent Topics

41 Topics
Spanner
Axelera Team
SpannerAxelera Team
posted in AI Trends and Innovations
Have you seen any projects combining CNC and AI?

About 40m of 2020 aluminium profile just arrived for the CNC machine I’m building! I suddenly feel like this might be a bigger job than I initially planned…Anyway, that’s for another day. What’s crossed my mind now is whether anyone’s seen any projects that combine CNC with vision AI? I’m thinking it’d be cool if I could combine the two, since I’m building one machine anyway.Predictive maintenance, maybe? Although this is a homemade machine, so that’s not a huge issue.Perhaps positioning and alignment could be something AI could get involved with? That could potentially be useful and reduce waste. Any other ideas of existing projects anyone’s seen in this area? 

70
Spanner
Axelera Team
Fabrizio
Axelera Team
FabrizioAxelera Team
Momentum and Milestones: Axelera AI’s Progress in a Changing WorldBlog

The world has been chaotic lately - market swings, tariffs, companies being acquired (Kinara by NXP), other companies are refusing acquisition (Furiosa allegedly declined an $800m buyout from Meta), DeepSeek made everyone question the future of closed AI models and model scaling, while OpenAI’s Sam Altman committed to an “open-weight” AI model to come out this summer - something no one thought would happen.As I was reflecting on all of these, and Axelera’s state in it all, I find myself incredibly grateful.Despite everything happening around us, our team has been focusing on what we can control: solving customer problems, building world-class technology, and partnering with some of the world’s best technology providers. So much to be thankful for, and I want to share some of these reasons with you.We have seen amazing progress towards our vision of bringing artificial intelligence to everyone, to truly democratize what could be the most revolutionary technology we have seen in our life

6783
Q
S
Cadet
Semih_Cadet
posted in The Axelera Forum
Legal Issues and Axelera

Hi everyone,I am Semih. I am a PhD researcher at University of Tuebingen on Open Source and AI. I am trying to understand Axelera’s hardware and in future I am planning to research legal and regulatory issues related to AI hardware, role of open source hardware, licensing contracts, regulatory compliance especially with Cyber Resilience Act, GDPR and AI Act. I was wondering if anyone is conducting a similar research or are there any pain points or interesting discussions under these themes?Thanks in advance

100
S
Cadet
P
Cadet
pautu227Cadet
asked in Metis M.2
Trying to setup rpi5 with metis m2

​@Victor Labian I’m also trying to setup rpi5 with metis m2. I have question what host OS you was using and have you changed anything in config.txt?

769
V
Axelera Team
F
Cadet
floritCadet
asked in Metis M.2
Metis M.2 on Raspberry Pi 5 - Driver Installation and Docker Setup Issues

hello I hijack this a little bit … once I type “lspci” it gave me following output on a pi 5 0001:00:00.0 PCI bridge: Broadcom Inc. and subsidiaries BCM2712 PCIe Bridge (rev 30)0001:01:00.0 Processing accelerators: Axelera AI Metis AIPU (rev 02)0002:00:00.0 PCI bridge: Broadcom Inc. and subsidiaries BCM2712 PCIe Bridge (rev 30)0002:01:00.0 Ethernet controller: Raspberry Pi Ltd RP1 PCIe 2.0 South Bridge how to activate the device and getting to work with my python3 script  

1310
F
Cadet
S
Cadet
SaitejaCadet
asked in Metis M.2
Deploying the YOLOPv2 model on to Metis M2 for lane detection

Hi, I am also working on same thing deploying the YOLOPv2 model on to Metis M2. I have downloadded the data set from bdd100k and arranged in this manner.voyager-sdk/├── data/│   └── yolopv2_dataset/│       ├── images/│       ├── labels/│       ├── cal.txt│       ├── val.txt│       └── data.yaml├── customers/│   └── my_yolopv2/│       └── yolopv2.pt└── yolopv2-custom.yamlso when i am trying to deploy the model using yaml I am getting this error(venv) aravind@aravind-H610M-H-V2:~/Desktop/voyager-sdk$ ./deploy.py customers/my_yolopv2/yolopv2.yamlINFO    : Using device metis-0:1:0INFO    : Detected Metis type as pcieINFO    : Compiling network yolopv2-custom /home/aravind/Desktop/voyager-sdk/customers/my_yolopv2/yolopv2.yamlINFO    : Compile model: yolopv2-customINFO    : Imported DataAdapter ObjDataAdaptor from /home/aravind/Desktop/voyager-sdk/ax_datasets/objdataadapter.py/home/aravind/.cache/axelera/venvs/93f45ae3/lib/python3.10/site-packages/torch/serialization.py:779: UserWarning: 'to

443
Spanner
Axelera Team
Z
Cadet
zefirCadet
asked in Voyager SDK
Is graphical video output expected to work from within the docker SDK?

Hi all,I’m a docker noob, so please bear with me if this is a stupid question.To get the SDK operational on my Ubuntu-24.04 system I followed the Install Voyager SDK in a Docker Container and can make it up to test-running some inferences. Alas, while I expected the videos to be displayed in a dedicated window at full resolution, for me they are shown in the active text-console (ncurses-based I assume), obviously coarse and slow.I tried to learn how to activate graphics output from within docker, but all I got was that docker essentially is meant to be text-only and graphics usually is provided via web-browser of pages served from within docker.So to put it bluntly: those videos showing Voyager SDK performing object detection at exhibitions at full-resolution and full-framerate, are those from within docker or are they run from a native ubuntu-22 system?If it is possible to tune docker to have that video output, a mini-howto would help a lot. 

172
Spanner
Axelera Team
Z
Cadet
zefirCadet
asked in Metis PCIe
Reset and resource allocation failures on Metis PCIe device on AMD64 system

Dear all,I am trying to get my hands on the Metis PCIe card on my current working PC, beingmainboard: Asus Prime X570-P CPU: AMD Ryzen 9 3900X OS: Ubuntu 24.04.2 LTS, 6.8.0-58-generic x86_64The first issue I face is that the warm-reset does not seem to work for me, i.e. if I issue a reboot without re-powering the system, the Metis device does not show up in lspci, while after a power-down (with physical PSU switching off) it is there. This is reproducible and based on my experience points to either my mainboard BIOS issuing a wrong reset sequence to the PCIe slot, or the device itself not issuing a full reset sequence on warm-boot. Since I know how to get the card detected, this is only an annoyance - but I saw that also another user of AMD64 system in the forum has similar issues and the problem might go deeper.Which leads to the second issue I am observing and which I filed already a support request but realized here would be the better place to discuss it.According to David M. to ch

8616
Spanner
Axelera Team
Spanner
Axelera Team
SpannerAxelera Team
asked in LLMs, NLP, Neural Networks
May the Source (Code) be With You!

May 4th is almost upon us, so naturally we’re thinking about AI… in space. More specifically: the AI and droid legends of the Star Wars universe.From multilingual protocol units to rogue assassin bots with surprising moral compasses, Star Wars has given us some of the most iconic (and occasionally chaotic) artificial intelligences in sci-fi.So here’s the question...

330
Spanner
Axelera Team
D
Cadet
DrouenCadet
asked in Metis M.2
Troubleshooting installation

Hello my name is Kevin, i am trying to work on a project that the m.2 is going to interfere with my cpu and later on maybe my gpu.is there a way, and another question whats the command to go on rtsp?  my system is:AMD 9950X3D CPUAMD 7900XTX GPU48 gb DDR5 on 8200MHZMetis Axelera AI on an m.2 port. 

6212
Dr Data
Ensign
Spanner
Axelera Team
SpannerAxelera Team
posted in AI Trends and Innovations
What Real-Time 8K Inference Looks Like

Check out our own ​@jonask-ai showing real-time, high-precision 8K object detection demo at Embedded World. An Aetina box, four Metis cards, and a whole lot of tile-based inference.What’s cool is it runs YOLOv8l across 380 overlapping tiles per second—no downscaling—at just 57% compute. And it’s not a one-off: same setup anyone can build with Voyager SDK + a YAML file.Noice!You can find out a bit more about this in ​@Doug’s blog post, too.

560
Spanner
Axelera Team
Ioannis
Axelera Team
IoannisAxelera Team
Voyager SDK is Available Now on GitHubBlog

The importance of software for unlocking the value of artificial intelligence cannot be overstated: the most powerful hardware in the world is just a paperweight without a usable software stack. At Axelera AI we’ve built the Voyager Software Development Kit (SDK) to give developers and ML engineers a simple solution for developing and deploying AI. Today, we’re excited to share that the Voyager SDK is now publicly available on our GitHub page.   IntroductionThe Voyager SDK is an end-to-end integrated software stack for Axelera AI’s inference platform which has been designed for performance, efficiency and ease of use. It enables developers to deploy pre-trained machine learning (ML) models, and to construct end-to-end optimized application pipelines quickly and easily. Whether you have trained your own model weights, are using an open source model with pre-trained weights, or want to build on one of the models offered in our model zoo, Voyager provides an effortless path to deploy and

94810
Spanner
Axelera Team
M
Cadet
MartianCadet
posted in Metis M.2
Testing and M.2 card on an AMD system, not recognized by lspci

This is a very interesting thread as I’m testing and M.2 card on an AMD system as well and am having what seems to be the exact same problems. Things added / tried: kernel command line includes md_iommu=off pcie_aspm=off update-pciids → resulted in card having the correct name in lspci : Axelera AI Metis AIPU (rev 02) axdevice --refresh -v results in: INFO:axelera.runtime.axdevice:Removing 0000:20:04.0INFO:axelera.runtime.axdevice:PCIE rescan0000:25:00.0 : Axelera AI Metis AIPU (rev 02)INFO:axelera.runtime:Found PCI device: 25:00.0 Processing accelerators: Axelera AI Metis AIPU (rev 02)INFO:axelera.runtime:Found AIPU driver: metis                  90112  0WARNING:axelera.runtime:4PCI device count mismatch: lspci=1, triton=0(removed the traceback) just tried rerunning install.sh which mentions at the end: building operators refreshing pcie and firmware WARNING: Failed to refresh pcie and firmware  tried removing metis-dkms / rmmod metis and reinstalling results in same warning as ab

634
D
Cadet
O
Ensign
oguzEnsign
asked in Metis M.2
Transforming the Attention mechanism into Convolutional

Hello,I know that axelera does not work in the attention mechanism. For example, can we convert the attention mechanism in the yolov11 architecture to convolution? Have you done any research or read on this subject? Does Softmax structure work in Axelera? (as far as I know, no) Is there a way to turn Softmax into a working structure?

572
B
Axelera Team
O
Ensign
oguzEnsign
asked in Metis M.2
How does Axelera speed up models?

Hello, is there a detailed document where I can learn about the Axelera chip architecture and how to do the quantization in it?Thanks

382
B
Axelera Team
T
Cadet
TheLinkedConnectionCadet
ai actors merged in video w/real actors?New

I'm developing a new way of video with or without AI. I am marketing both small & large companies using their people as limited actors. I saw Titanium and i was very curious, can I merge dynamic ai actors into video with real actors?  I am getting paperwork together to start, so I have time to learn. My background is in 2 & 3D animation & graphic design (degrees in both bs & bfa) What is the learning curve?  

1
151
Spanner
Axelera Team
S
Ensign
saravananEnsign
asked in The Axelera Forum
Deploying lane dection model (yolopv2.onnx) on metis M2

Hello,I am currently working on lane detection model (YOLOv2) which is in .pt format I have converted to ONNX format.Here are the terminal responses I have got..root@aetina:/voyager-sdk# cd YOLOPv2root@aetina:/voyager-sdk/YOLOPv2# lsLICENSE    data     demo2.py        models            runs   yolopv2-envREADME.md  demo.py  export-onnx.py  requirements.txt  utilsroot@aetina:/voyager-sdk/YOLOPv2# cd modelsroot@aetina:/voyager-sdk/YOLOPv2/models# ls__pycache__  common.py  experimental.py  yolopv2.onnx  yolopv2.ptroot@aetina:/voyager-sdk/YOLOPv2/models# I am having the ONNX format of the model so now I want to deploy the model onto Metis M2. When I am trying to compile the yolopv2.onnx model but i am encountering this issue.root@aetina:/voyager-sdk# /voyager-sdk/bin/voyager-compile models/yolopv2.onnx --output-dir build/yolopv2 --input-shape "1,3,640,640"bash: /voyager-sdk/bin/voyager-compile: No such file or directoryI request to guide me to successfully deploy the model.Note; I am using

494
S
Ensign
quant.geek
Cadet
quant.geekCadet
asked in Voyager SDK
Ubuntu 24.04 support, and Docker containers

Seems Ubuntu 24.04 is not supported at the moment. What are the plans for this version? Can you give some insight on what needs to be changed to get this installed on that version?

18921
quant.geek
Cadet
Z
Cadet
zefirCadet
asked in Metis PCIe
Flashing PCIe Metis with the latest FW version

After finally getting my metis card properly accessible through PCI, it turned out I need to upgrade the FW to the latest version:(venv) root@zefir-PC:~/voyager-sdk# axdevice [libdmabuf.c:1720] Device command returned an error code.[libdmabuf.c:1721] Consider inspecting the device log via `triton_trace --slog`.[...]Device 0: metis-0:a:0 board_type=pcie fwver='v1.2.0-rc2+bl1-stage0' clock=800MHz(0-3:0MHz) mvm=0-3:-1%(venv) root@zefir-PC:~/voyager-sdk# triton_trace --slogVersion mismatch!Actual="v1.2.0-rc2+bl1" Expected="v1.2.5"Please update device firmware before continuing.I proceeded like described in FW flashing tutorial at github(venv) root@zefir-PC:~/Axelera.ai/firmware_release_public_v1.2.0# ./flash_update.sh flash_bundle.img>>> loading new AIPU runtime stage0>>> the following operations updates the flash content Make sure to only run this operation if really needed Also make sure to not power cycle the devices or it might be rendered non-functional.do y

324
Spanner
Axelera Team
Spanner
Axelera Team
SpannerAxelera Team
posted in LLMs, NLP, Neural Networks
Built-In AI or Subscription-Based?

This new survey suggests people (Apple fans, that is) are very open to the idea of paying a subscription for “Apple Intelligence” features. Which raises the question as to whether AI tools are going to become just another line on your monthly bill.We’ve moved quickly from “AI is cool” to “you might need it just to keep up.” And if it’s locked behind paywalls? That could become a problem once it’s a necessity rather than a nice-to-have.Curious what you all think – will AI be a must-have, or a paid perk for you?https://www.macrumors.com/2025/04/22/apple-intelligence-payment-survey/

70
Spanner
Axelera Team
Z
Cadet
zefirCadet
asked in Metis PCIe
Metis PCIe card not detected after reboot

Hi all,not sure if my issue is related to this one, but symptoms are at least similar, so I hope it matches best here. Otherwise sorry for derailing.I am trying to get my hands on the PCIe metis card on my AMD64 based system driven by an ASUS PRIME X570-P mainboard under Ubuntu-24.04 (6.8.0-58-generic x86_64). First issue was that the card did no show up under lspci, no matter which kernel boot options I tried. Then I looked into the mainboard BIOS to double-check the peripheral config and noticed that the PCIe slots are not fully independent. In fact, the board has two PCIex16 slots and three PCIex1, of whose the latter are meant for expansion cards and the former for graphic cards. They are pre-configured for ‘auto’ mode, which means that the BIOS detects how the slots are populated and splits available lanes between the two x16 slots.I changed this ‘auto’ setting to manually split the lanes to ‘x8+x8’ and after rebooting the card is now visible to the kernel:zefir@zefir-PC:~$ lspci

373
Spanner
Axelera Team
J
Axelera Team
JonasAxelera Team
🎥 Demo Drop: Multimodal segmentation and detection on Metis with Fast Segment AnythingUnder Consideration

This short video showcases an early experiment using the Voyager SDK and Metis AIPU to run Fast Segment Anything (FastSAM). It is based on YOLOv8-seg generating masks and CLIP to link the text prompt with visual masks in the video stream.What’s exciting is the flexibility—no need for retraining on specific classes. Just change the text prompt and it finds the object. This is still a test setup, but I thought it might spark a few ideas and show what’s becoming possible on the edge with Metis. Keen to hear what you guys think, or how you’d use this!

11
994
B
Axelera Team
Axelera CMX
Axelera Team
Axelera CMXAxelera Team
posted in The Axelera Forum
We have our first Ensigns! 🥇🥈🥉 Report to the bridge! 📯

This is a quick shoutout out to ​@ChinmayMulgund and ​@oguz for becoming the first members of Axelera AI’s community to earn the rank of Ensign! And practially at the same time (close enough that I can’t tell who got there first).Every activity on the community earns you points, and helps you climb the ranks. The opening rank is a Cadet, which you receive after your first action on the community. But it takes effort to climb the ranks, so a big congratulations to our first two Ensigns for getting there so quickly!Who’s going to be the first Lieutenant, I wonder? 

220
Axelera CMX
Axelera Team
C
Ensign
ChinmayMulgundEnsign
asked in Metis M.2
Guidance on Running Lane Detection Model (culane_res18.onnx) on Metis M2

Hello,I am currently working with a lane detection model named culane_res18.onnx, and I would like to know how I can run this model on the Metis M2 device.Here are some details about the setup:The model is culane_res18.onnx, which is a lane detection model. The training dataset is from the CULane dataset, and I have the link to the dataset here. The model file can be found here.Could you kindly guide me on how to run this lane detection model on Metis M2? I would appreciate any step-by-step instructions or basic guidance on how to deploy this model, considering that I have both the ONNX model and the dataset.Looking forward to your response!Best regards,Chinmay Mulgund

16713
C
Ensign
C
Ensign
ChinmayMulgundEnsign
asked in Voyager SDK
Issues Deploying and Running yolov5m6.onnx on M.2

I am encountering issues when trying to deploy and run the YOLOv5m6 ONNX model on the Metis M.2. Below are the details of the steps I followed and the issues that arose:1. Deployment Process:I used the following command to deploy the model:./deploy.py customers/mymodels/yolov5m6-object-detection.yamlThe deployment was successful, but I encountered warnings about memory overflow during the buffer fitting process:LowerTIR failed to fit buffers into memory after iteration 0/4. Pool usage: {L1: alloc:4,195,840B avail:4,194,304B over:1,536B util:100.04%, L2: alloc:6,864,384B avail:8,077,312B over:0B util:84.98%, DDR: alloc:54,275,328B avail:260,046,848B over:0B util:20.87%}2. Inference Process:I then attempted to run inference with the following command:./inference.py c-yolov5m6-object-detection chinmay/Vehicle-CV-ADAS/assets/VID-20250208-WA0008.mp4 --save-output ~/Desktop/voyager-sdk/chinmay/saved_videos/output.mp4The model failed to run due to a ShapeInferenceError related to incompatibl

311
Spanner
Axelera Team

Badge Winners

Show all badges
Code of Conduct

Address

HTC5, High Tech Campus
5656 AE Eindhoven
The Netherlands
Email: info@axelera.ai

Menu

AI Accelerators

AI Software

Metis Evaluation Systems

Partners

Product Inquiry

Company

Our Story

Our Team

Careers

News

Contact

Reducing CO₂ with
Axelera’s Forest

Follow Us

LinkedIn

Twitter

Sign Up for Our Newsletter

EU flag Axelera AI is co-funded by the European Union’s Horizon research and innovation programme.

Privacy Policy  |  Cookie Policy  |  Terms of Sale  |  Vulnerability Disclosure Policy

Copyright 2025 © All rights Reserved.