home.social

#jetson — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #jetson, aggregated by home.social.

fetched live
  1. Banana Pi BPI-CM5 Pro is a computing module powered by the Rockchip RK3576, 6 TOPS computing power NPU,8-32G RAM and 8-128G eMMC,are you want o do Edge AI project or replace Raspberry Pi CM4?
    docs.banana-pi.org/en/BPI-CM5_
    #NVIDIA #Jetson #Nano #BananaPi #RaspberryPi #AI #Robots #Edge

  2. Banana Pi BPI-AIM7 with Rockchip RK3588, 6 TOPS computing power NPU,8-32G RAM and 8-128G eMMC,what Edge AI project are you want to do.
    docs.banana-pi.org/en/BPI-AIM7
    #NVIDIA #Jetson #Nano #BananaPi #RaspberryPi #AI #Robot

  3. Love to meet the "genius" at Nvidia who decided that the version of the boot loader firmware in QSPI needs to match the launcher image on the system disk exactly or the system won't boot. Makes upgrading so much more complex for no benefit. Just lazy programming allowing them to not care about backwards compatibility. #FAIL #Nvidia #Jetson

  4. K3s на колёсах: HA-кластер для автопарка с ARM-агентами, WireGuard и офлайн-режимом

    Большинство статей про Kubernetes на границе (edge) описывают более-менее стабильную топологию: несколько узлов в одном или нескольких дата-центрах, предсказуемая сеть между ними, редкие сетевые партиции как исключение, а не норма. У нас всё наоборот: control plane живёт в офисе/датацентре, а agent-узлы физически стоят в транспортных средствах автопарка — едут по городу, паркуются в подземных гаражах без связи, теряют VPN на десятки минут за смену. Разрыв связи с control plane — это не инцидент, это штатный режим работы системы, который нужно было спроектировать с самого начала, а не «подкрутить» постфактум. В этой статье — то, как устроен наш K3s-кластер для автопарка: HA control plane на kube-vip, транспорт через WireGuard поверх MikroTik, разнородный флот из amd64- и ARM-агентов (включая Jetson), и отдельно — самая интересная часть: как мы подступаемся к вопросу «может ли под на границе продолжать работать, когда control plane временно недостижим».

    habr.com/ru/articles/1079024/

    #Kubernetes #K3s #DevOps #WireGuard #kubevip #Jetson #Edgecomputing #cgroups

  5. NVIDIA has announced a new entry-level edge AI board, the Jetson Orin Nano 2. The upgraded Nano is not only twice as fast as its predecessor, but unexpectedly, features an outright new Orin SoC at its heart#Ampere #Edge #Jetson #NVIDIA #Orin
    NVIDIA Announces Jetson Orin Nano 2: Entry-Level Edge Board Gets New Ampere Silicon
  6. Jim Carrey vai voltar às telonas como protagonista de um filme live-action dos Jetsons, anunciou a Warner Bros. nesta quarta-feira. O ator, conhecido por Ace Ventura e Sonic, viverá George Jetson em uma adaptação da clássica série de ficção científica.

    Este será o primeiro projeto do carismático comediante fora da franquia Sonic desde 2016, marcando seu retorno após um período de baixa perfil... #Jetson

  7. Now, if you didn’t know NVIDIA made ARM single board computers, I can’t really blame you. When most people think of TeamGreen™ they think GeForce, CUDA, and leather jackets.

    interfacinglinux.com/2026/08/0 #ARM #JETPACK #JETSON #LLM #NVIDIA #reComputer #REVIEW
  8. Can #Copilot and I replace the GNOME desktop GUI of Ubuntu on my #Jetson Orin Nano with our own Single Window AEye GUI? No more silly Linux #GNOME, #XFCE, #MINT, ... desktop GUI. We use Openbox and Qt6 for this project.

    Our AEye GUI will have a single window with a tab system and a collapsible AI panel. No more task bar, sidebar, ... All these Linux desktop environments are inefficient and are eating up all the resources of #Ubuntu #Linux, which is already inefficient and bloated.

    #AI #linux

  9. #Jetson Orin Nano is unfriendly as an embedded board for simple robotics or #agenticAI. You can't get cheap audio using cheap DAC. Unlike the GPIO pins on any #esp32 board that have fixed destinations, the pins on the Jetson board vary depending on your Jetpack version.

    On Jetpack 7.2, jetson-io.py configuration is ignored by bootloader. Changing extlinux.conf and dtbo does not work. I burned my PCM5100 DAC board after a wrong pin connection to the fake 40 pin GPIO. HDMI is not cheap!

    #AI

  10. #Jetson Orin Nano #Ubuntu is very unfriendly with no good native apps and is relatively slow running most programs. So, it is only good for doing some small LLM inferencing and #edgeAI stuff like #LiDAR which I don't care cause I will be using #esp32 for that purpose.

    But the good news is, I can use Pi Apps which uses #Flatpak, allowing some of my favorite apps. like #Obsidian notes, to run on this Jetson board that runs on an ARM 64 version of Ubuntu.

    #AI #LLM

  11. The Jetson Orin Nano board is not really that cheap. $249 includes only the mother board with no storage. That 8GB RAM is shared by the CPU and the GPU. But the Ubuntu Linux distro used over 2G RAM leaving only 6G for the GPU and other apps. The Chromium browser even has problems loading some websites, like Bestbuy, that is heavy on javascripts. The Wifi seems weaker than my esp32 p4 c6.

    #Jetson Orin Nano is only good at running 1 - 3B small #LLMs. It gives 19 tokens/s on a 3B Q4.

    #AI #LLM

  12. After struggling for a couple of hours, my new #Jetson Orin Nano is finally alive. There are so many issues in this supposedly simple installation process. 1. I had to find that fc_rec pin and connect it to GND using a jumper. This silly pin is hiding under the CPU/GPU and the fan. 2. The data transfer usb cable must be high grade.

    This cheap card is not cheap at all. I had to use my 1T WD NvMe card cause it has zero memory out of the box. That alone costs me $140.

    #Nvidia #AI

  13. NVIDIA has announced the addition of two new mid-range boards to their Jetson Thor lineup, the T3000 and T2000, which will arrive in Q1'27. The new boards aim to be a cheaper offering for customers feeling pressured by the high cost of memory#Edge #Jetson #NVIDIA #Thor
    NVIDIA Announces Expanded Jetson Thor Lineup with Mid-Range T3000 and T2000 Modules
  14. The Robotic Realm of Open Duck Mini: Bringing Gemma 4 to Life Welcome to the fascinating world of Open Duck Mini v2 robots, a project that marries technology and creativity to breathe life into rob...

    #Google #Robotics #ct: #AIG; #gds:N/A; #Gemma #4 #Jetson #Orin #Nano #pr_pr:

    Origin | Interest | Match
  15. From Budget Microcontrollers to High-End AI Boards: Navigating the Robotics Hardware Spectrum The year 2026 presents a vibrant tapestry of possibilities in robotics, from the humble $10 microcontro...

    #Maker #Robotics #Microcontrollers #NVIDIA #Jetson #physical #AI

    Origin | Interest | Match
  16. Планирование движения для ровера на ходовой Ackerman'а

    Проверяем, может ли GPU-планировщик MPPI-Generic работать без глобального планера. Самодельный симулятор, контроллер с приоритетом угловой скорости, сравнение с State Lattice и RPP vs MPPI, справился, но есть нюансы.

    habr.com/ru/articles/1039784/

    #ros #ros2 #diy #jetson #cuda #робототехника

  17. Sipeed Crams 32GB LPDDR5 and a 60 TOPS NPU Into a Compact RISC-V Board That Hits 15 Tokens/s on Qwen-3.5 35B AI LLMs Sipeed has unveiled its new RISC-V powered SBC platform, the K3 series, which ca...

    #Featured #News #Sticky #Bianbu #OS #Jetson #Orin #K3 #Key #Stone #K3

    Origin | Interest | Match
  18. Running DeepSeek R1 on Raspberry Pi, Jetson, and MacBook: A Journey in Local AI Execution In the dynamic world of artificial intelligence, running models locally offers both freedom and privacy. I ...

    #AI #Large #Language #Models #Jetson #Orin #Nano

    Origin | Interest | Match
  19. I spent the weekend nerding out on my Jetson Nano.
    I am working on depth estimation with stereo vision and car detection using YOLO.
    The pipeline is a bit slow at the moment, but it is getting there !

    #nvidia #jetson #artificialintelligence #DIY #raspberrypi

  20. 🚀 Chạy thành công mô hình 14B trên 3 Jetson Orin Nano! Tốc độ 3-4 token/giây qua mạng, nhưng đã tối ưu code Python hiệu quả. #AI #LLaMA #Jetson #DeepLearning #CôngNghệ

    reddit.com/r/LocalLLaMA/commen

  21. :boost_requested: boosts appreciated

    genuinely, nvidia pisses me off SO much that I am willing to entirely switch platforms from this shit. is there any alternative to a nvidia jetson for using with ros for robotics-type applications? (example) I need hardware acceleration because I'm going to be:

    • doing SLAM on LiDAR data & several video streams (using RTAB-Map)
    • running some ML models on the video streams
    • hardware encoding of video streams (some of this has already been offloaded to dedicated devices, however I'm still going to need to do it for some other streams which don't use those devices)

    willing to use something that isn't as dedicated to open source, so long as it's reasonably supported.
    bonus points for open source, though

    misc tags for visibility:
    #nvidia #jetson #nvidiajetson #ros #ros2 #robotics #openrobotics

  22. We almost bricked a Nvidia Jetson Orin Nano by selecting ACPI mode #ctrlos #jetson #nix #nixos #nvidia

Share on Mastodon

Enter the server where you have an account.