Artificial Intelligence Computing Leadership from NVIDIA
hp proliant dl380p gen8 - nvidia tesla m40 not working
I`m tring to install nvidia tesla m40 card into hp proliant dl380p gen8 p70 server. its not working, found below error in DMESG log . [ 959.113417] NVRM: The system BIOS may have misconfigured your GPU. [ 959.113461] nvidia: probe of 0000:24:00.0 failed with error -1 [ 959.113482] NVRM: The NVIDIA probe routine failed for 1 device(s). [ 959.113483] NVRM: None of the NVIDIA graphics adapters were initialized! [ 959.113564] nvidia-nvlink: Unregistered the Nvlink Core, major device number 245 [ 959.188166] nvidia-nvlink: Nvlink Core is being initialized, major device number 245 [ 959.189044] NVRM: This PCI I/O region assigned to your NVIDIA device is invalid: NVRM: BAR1 is 0M @ 0x0 (PCI:0000:24:00.0) Server OS: ubuntu 16.04 LTS Please somenone help me resolve this.. Thanks in advance
I`m tring to install nvidia tesla m40 card into hp proliant dl380p gen8 p70 server. its not working,

found below error in DMESG log .

[ 959.113417] NVRM: The system BIOS may have misconfigured your GPU.
[ 959.113461] nvidia: probe of 0000:24:00.0 failed with error -1
[ 959.113482] NVRM: The NVIDIA probe routine failed for 1 device(s).
[ 959.113483] NVRM: None of the NVIDIA graphics adapters were initialized!
[ 959.113564] nvidia-nvlink: Unregistered the Nvlink Core, major device number 245
[ 959.188166] nvidia-nvlink: Nvlink Core is being initialized, major device number 245
[ 959.189044] NVRM: This PCI I/O region assigned to your NVIDIA device is invalid:
NVRM: BAR1 is 0M @ 0x0 (PCI:0000:24:00.0)

Server OS: ubuntu 16.04 LTS

Please somenone help me resolve this..

Thanks in advance

#1
Posted 03/16/2018 07:21 AM   
[quote=""][ 959.189044] NVRM: This PCI I/O region assigned to your NVIDIA device is invalid: NVRM: BAR1 is 0M @ 0x0 (PCI:0000:24:00.0) [/quote] Try to reconfigure PCIe (mem BAR) address space (MMIO) settings in RBSU/BIOS (something like "MMIO above 4G" or "Above 4G Decoding", or "MMIO base" set to lower value) or remove/disable other unnecessary add-on cards and use 64bit OS. HPE RBSU/BIOS have problems with that (see similar problems [url]https://gridforums.nvidia.com/default/topic/298/[/url]). To check address space requirements use "lspci -s 24:00.0 -vv | grep Region". PS: Tesla M40 is for "Deep Learning"/CUDA processing not for GPU/vGPU and you can check more relevant forums [url]https://devtalk.nvidia.com/[/url].
said:[ 959.189044] NVRM: This PCI I/O region assigned to your NVIDIA device is invalid: NVRM: BAR1 is 0M @ 0x0 (PCI:0000:24:00.0)

Try to reconfigure PCIe (mem BAR) address space (MMIO) settings in RBSU/BIOS (something like "MMIO above 4G" or "Above 4G Decoding", or "MMIO base" set to lower value) or remove/disable other unnecessary add-on cards and use 64bit OS. HPE RBSU/BIOS have problems with that (see similar problems https://gridforums.nvidia.com/default/topic/298/). To check address space requirements use "lspci -s 24:00.0 -vv | grep Region".
PS: Tesla M40 is for "Deep Learning"/CUDA processing not for GPU/vGPU and you can check more relevant forums https://devtalk.nvidia.com/.

#2
Posted 03/16/2018 08:10 AM   
Scroll To Top

Add Reply