Skip to main content

Hello everyone,

I’m new to working with this hardware and recently installed the Metis M.2 chip (Firmware version: v1.2.0-rc2+bl1-stage0) on an ASRock Z170 Pro4S motherboard. The device was detected successfully.

I’ve installed the Voyager SDK and attempted to run the following command from the Quick Start Guide:

./inference.py yolov5s-v7-coco dataset --no-display

However, I encountered an input/output error when executing the command. I’ve included the full error message below for reference.

I would appreciate any guidance or suggestions you may have to help resolve this issue.

Thank you in advance for your support.

 

INFO    : Using default val dataset INFO    : Using device metis-0:5:0 INFO    : Network type: NetworkType.SINGLE_MODEL INFO    : Input INFO    :   └─detections INFO    : Imported DataAdapter ObjDataAdaptor from /home/tud/voyager-sdk/ax_datasets/objdataadapter.py Stream Playing:  71%|██████████████████████▏        | 5/7 00:00<00:00,  6.26/s]AxeleraDmaBuf.cpp:234] UIO_IOCTL_DMA_XFER failed: Input/output error .ERROR]AaxeShareMemoryExecute]: Dmabuf transfer failed. ERROR]oaxeCommandQueueExecuteCommandListsSync]: Level-zero memory operation failed: 0x70010001. terminate called after throwing an instance of 'std::runtime_error'   what():  axr_load_model_instance failed : Error at zeCommandQueueExecuteCommandLists(cmdqueue, n_cmdlists, cmdlists, nullptr): cmdqueue_run_cmdlists: 309: Exit with error code: 0x70010001 : ZE_RESULT_ERROR_NOT_AVAILABLE Aborted (core dumped)

Hi ​@sifat0433! Welcome to the community!

On some x86 systems it’s worth checking that IOMMU (VT-d) and ACS (Access Control Services) are enabled in your BIOS settings.

Could you check the following BIOS options?
    •    VT-d or Intel Virtualisation for Directed I/O: Make sure it’s enabled.
    •    ACS Control: If available, enable this too.
    •    Above 4G decoding: Also enable this if it’s an option—it helps with memory allocation for PCIe devices.

After making changes, reboot and try your inference command again. Let me know how it goes, and we can dig deeper if needed. 👍


Hi. Thank you so much. 

Everything was enabled except 4G decoding. I enabled it. It worked twice with and without display at 10 % utilization. Now again I have the same issue. Also, after few attemps the m.2 link gets down. Rescan or refresh does not work. I have to turn off and on the PC and need to reload the firmware. 


I’ve done some preliminary investigation and would like to know the recommended operating temperature range for the Metis M.2 chip. I suspect the issue may be thermal-related, as the card did not include a heatsink or active cooling solution.


I’ve done some preliminary investigation and would like to know the recommended operating temperature range for the Metis M.2 chip. I suspect the issue may be thermal-related, as the card did not include a heatsink or active cooling solution.

Did that come from the Axelera web store, ​@sifat0433?


Yes. My university ordered the metis m.2 card from the Axelera shop directly. As far as I remember it was a pre-order.


Hi,

I’m continuing to troubleshoot an issue with the Metis M.2 card. I have installed a heatsink to address potential thermal problems, but unfortunately, the same error persists.

I also ran the following commands as suggested in the documentation:

axdevice --reload-firmware  
axdevice --pcie-rescan

However, the issue remains unresolved. When I attempt to run inference, I receive the following error in the system log:

e  367.739896] axl 0000:03:00.0: DMA error RD CH0 (ctrl 0x2)

 


I'm happy to share that the issue has been resolved.

The main problems were related to thermal management and M.2 slot communication. I resolved them by using an M.2 to PCIe adapter equipped with a heatsink. After this setup change, the system is now functioning as expected, and the Metis M.2 card runs inference tasks without errors.


I'm happy to share that the issue has been resolved.

The main problems were related to thermal management and M.2 slot communication. I resolved them by using an M.2 to PCIe adapter equipped with a heatsink. After this setup change, the system is now functioning as expected, and the Metis M.2 card runs inference tasks without errors.

Ah, that’s really good to know, thanks for updating ​@sifat0433! Glad it’s up and running - good work on figuring that out.

I’d be interested in seeing a photo of your new setup, if you get chance!


I'm happy to share that the issue has been resolved.

The main problems were related to thermal management and M.2 slot communication. I resolved them by using an M.2 to PCIe adapter equipped with a heatsink. After this setup change, the system is now functioning as expected, and the Metis M.2 card runs inference tasks without errors.

Ah, that’s really good to know, thanks for updating ​@sifat0433! Glad it’s up and running - good work on figuring that out.

I’d be interested in seeing a photo of your new setup, if you get chance!

 

Thank you. Sure. Here is the setup: 

 


That’s awesome, thanks ​@sifat0433! Really interesting to see that setup - it’s not something we’ve tried very much ourselves, using a PCIe adapter, so it’s great to see it working. Nice one!


Reply