Discussions

I plan on running 3 RK1 32GB and 1 Nvidia Jetson Orin for AI (hardware on the way). What is the optimal board install for the 4 cards? Should the Jetson go in slot 1 for HDMI etc? I'm planning on running some AWS localstack things and messing around with AI.

Hi, I am running an RK1 with my TuringPi2 v2.4. On RK1 I am running the Ubuntu 22.04 official server image. With method 2 of guide <https://docs.turingpi.com/docs/turing-rk1-flashing-os#running-off-an-external-block-device---method-2> I changed to my NVMe drive as you can see in the following: ubuntu@ubuntu:~$ lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS mmcblk0 179:0 0 29.1G 0 disk ├─mmcblk0p1 179:1 0 512M 0 part /boot/firmware └─mmcblk0p2 179:2 0 28.6G 0 part / mmcblk0boot0 179:32 0 4M 1 disk mmcblk0boot1 179:64 0 4M 1 disk nvme0n1 259:0 0 3.6T 0 disk ├─nvme0n1p1 259:1 0 512M 0 part └─nvme0n1p2 259:2 0 3.6T 0 part ubuntu@ubuntu:~$ sudo ubuntu-rockchip-install /dev/nvme0n1 This script will install the currently running system onto /dev/nvme0n1. It may take up to a few minutes - Continue? [y/N] y Creating partition table for /dev/nvme0n1. Creating filesystem on partitions /dev/nvme0n1p1 and /dev/nvme0n1p2. Counting files... Transferring 920 files (206 MB) from the bootfs to /dev/nvme0n1p1. Please wait! Cleaning up... Counting files... Transferring 78575 files (3753 MB) from the rootfs to /dev/nvme0n1p2. Please wait! Cleaning up... Done! ubuntu@ubuntu:~$ lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS mmcblk0 179:0 0 29.1G 0 disk ├─mmcblk0p1 179:1 0 512M 0 part └─mmcblk0p2 179:2 0 28.6G 0 part mmcblk0boot0 179:32 0 4M 1 disk mmcblk0boot1 179:64 0 4M 1 disk nvme0n1 259:0 0 3.6T 0 disk ├─nvme0n1p1 259:1 0 512M 0 part /boot/firmware └─nvme0n1p2 259:2 0 3.6T 0 part / After powering off, it seems this is not persistent: ubuntu@ubuntu:~$ lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS mmcblk0 179:0 0 29.1G 0 disk ├─mmcblk0p1 179:1 0 512M 0 part /boot/firmware └─mmcblk0p2 179:2 0 28.6G 0 part / mmcblk0boot0 179:32 0 4M 1 disk mmcblk0boot1 179:64 0 4M 1 disk nvme0n1 259:0 0 3.6T 0 disk ├─nvme0n1p1 259:1 0 512M 0 part └─nvme0n1p2 259:2 0 3.6T 0 part Did I miss any step? How can I make this persistent? Thanks, Luke

Can I use a Gen 4x4 or only Gen 3x4? Any card recommendations at 1TB please?

Hello Turing Pi Community, I am setting up an INA219 current sensor to monitor the power consumption of the compute modules on my Turing Pi 2 board. I need to know which power cables from the ATX power supply are actually powering the modules to connect my sensor correctly. Despite disconnecting the red +5V cables without affecting module operations, I'm unsure if the primary power source might be the +12V or +3.3V lines. Could someone confirm which power lines are used for the compute modules? Also, are there any precautions or specific configurations I should consider for this measurement setup? Thank you for any insights or advice!

If you've pre-ordered a Turing Pi 2.5, will you receive that version of hardware exactly, or will an updated version such as 2.5.1 ship with bugfixes instead?

Hi there, I have a Turing Pi 2 that I got from backing on Kickstarter. I have 2x 8GB RK1s as nodes 1 and 2 (which remain off at the moment as it looks like I need to get a different USB cable to flash them with rkdeveloptool \[as here](<https://docs.turingpi.com/discuss/65b024aa66172c003199e612>)) and 2x 8GB CM4 Lite (CM4008000) as nodes 3 and 4. I've flashed this \[Talos OS 1.7.0](<https://factory.talos.dev/image/ee21ef4a5ef808a9b7484cc0dda0f25075021691c8c09a276591eedb638ea1f9/v1.7.0/metal-arm64.iso>) onto SD cards for the 2x CM4s. Node 4 boots up fine. I can see it boot successfully using picocom in ssh, and I can apply a config with talosctl. However, I'm unable to get node 3 to boot. This is what I see on the serial terminal: > <br> > > U-Boot 2024.01 (Mar 01 2024 - 17:26:00 +0000) > > DRAM: 992 MiB (effective 7.9 GiB) > RPI Compute Module 4 (0xd03141) > Core: 207 devices, 16 uclasses, devicetree: board > MMC: mmc@7e340000: 0 > Loading Environment from FAT... Unable to read "uboot.env" from mmc0:1... > In: serial,usbkbd > Out: serial,vidconsole > Err: serial,vidconsole > Net: eth0: ethernet@7d580000 > PCIe BRCM: link up, 5.0 Gbps x1 (SSC) > PCI: Failed autoconfig bar 10 > PCI: Failed autoconfig bar 14 > PCI: Failed autoconfig bar 18 > PCI: Failed autoconfig bar 1c > PCI: Failed autoconfig bar 20 > starting USB... > No working controllers found > Hit any key to stop autoboot: 0 > \*\* Booting bootflow '[[email protected]_1](mailto:[email protected]_1)' with efi > Booting /efi\\boot\\bootaa64.efi I then actually get to GRUB, but no matter which boot option I choose, it then pretty quickly falls back to uboot: > EFI stub: Booting Linux Kernel... > EFI stub: Loaded initrd from LINUX_EFI_INITRD_MEDIA_GUID device path > EFI stub: Using DTB from configuration table > EFI stub: xiting boot services... > > U-Boot 2024.01 (Mar 01 2024 - 17:26:00 +0000) I notice that on node 4, I don't get those 5 failures about `PCI: Failed autoconfig bar xx`. Things I've tried so far: - I've tried swapping the nodes, swapping the SD cards, flashing a new SD card for node 3, and using a brand new spare carrier board. - I also have an Ethernet device that is passing through the Turing's internal switch to my router, so have tried unplugging that too. - I already had firmware 2.0.5, but I've tried applying the upgrade again too. Do you have any ideas on things to try?

For a board like this I would expect each node to have a dedicated 10Gbps and either a 2.5Gbps or 10Gbps port for management. 10Gbps is so cheap now and I'm actually moving my home network to 25Gpbs. I was really hoping the v2.5 would improve on this as I would really love to populate several of these boards with Orion NX modules for a Blender render farm. Unfortunately that bandwidth limitation makes this not a feasible solution for that use case. I'm not even sure what use cases 2x1Gbps is logical. <br>

Rasoi released new cm3 compatible Compute Module 4S. I have a yet unused Turing pi 1. Any reasons why the new modules would not work? Would be great if I can use them.

Ive been trying: - curl 'http\://turing/api/bmc/authenticate' -X POST -d '{"username":"root", "password":"turing"}' -k - ssh root@turing root@turing's password: Permission denied, please try again. root@turing's password: Permission denied, please try again. It's like my board doesn't have the password from the documentation (<https://docs.turingpi.com/docs/turing-pi2-bmc-ui>)... I flashed an SD card using the Raspberry Pi Imager 1.8.5 while flashing the TP2-Firmware-SDCARD-v2.0.5.img on it, the board doesn't seems to boot nor upgrade. I've changed two settings to have the SSH (<https://imgur.com/a/TnN4ab1>). The only thing that show up (without the SD card in it) is the old BMC user interface. I don't know what i missed in the documentation!

Are the LEDs for each slot on Turing Pi 1 boards connected to a GPIO? I was wondering if I could blink them, e.g. to determine which hostname is in which slot. Thanks.