Bdtyktl

I have a Thinkpad W520 with NVIDIA Quadro 1000M + integrated Intel GPU. I have just made a fresh install of Ubuntu 18.10 on it.

I want to achieve the following:

• working CUDA,


• working external display through DisplayPort/VGA,


• in Optimus mode, if possible (selected that in the BIOS).

Status before installing anything

I had nouveau set up by default. It finds the external monitor through both DisplayPort and VGA (the connectors on the laptop), but it does not work consistently: the external monitor gets switched off after some time.

What I've done

I tried to install CUDA along with the corresponding NVIDIA driver:

sudo apt install nvidia-cuda-toolkit

sudo ubuntu-drivers autoinstall

Had a mysterious crash in the middle of compiling the kernel modules, then finished the install after a reboot with dkpg --configure -a. In the end, it installed CUDA 9.1.

At this point, the laptop:

• boots in Optimus mode,


• it does not detect external VGA monitor (will need to test HDMI),


• 
  glxinfo says that direct rendering is Yes,


• the graphics is however not smooth (I see triangle-like artifacts appearing on screen when I move a window, or part of the screen is refreshed),


• after unpacking and compiling some samples from the original 9.1 toolkit installer to verify CUDA, it turned out that CUDA also has problems.

matrixMul does not work, see output:

Error! Matrix[204798]=0.00000000, ref=3.19999981 error term is > 1.000000E-06

Error! Matrix[204799]=0.00000000, ref=3.19999981 error term is > 1.000000E-06

Result = FAIL

Strange enough, deviceQuery does work, see output:

./deviceQuery Starting...



 CUDA Device Query (Runtime API) version (CUDART static linking)



Detected 1 CUDA Capable device(s)



Device 0: "Quadro 1000M"

  CUDA Driver Version / Runtime Version          9.1 / 9.1

  CUDA Capability Major/Minor version number:    2.1

  Total amount of global memory:                 1985 MBytes (2081751040 bytes)

MapSMtoCores for SM 2.1 is undefined.  Default to use 64 Cores/SM

MapSMtoCores for SM 2.1 is undefined.  Default to use 64 Cores/SM

  ( 2) Multiprocessors, ( 64) CUDA Cores/MP:     128 CUDA Cores

  GPU Max Clock rate:                            1400 MHz (1.40 GHz)

  Memory Clock rate:                             900 Mhz

  Memory Bus Width:                              128-bit

  L2 Cache Size:                                 131072 bytes

  Maximum Texture Dimension Size (x,y,z)         1D=(65536), 2D=(65536, 65535), 3D=(2048, 2048, 2048)

  Maximum Layered 1D Texture Size, (num) layers  1D=(16384), 2048 layers

  Maximum Layered 2D Texture Size, (num) layers  2D=(16384, 16384), 2048 layers

  Total amount of constant memory:               65536 bytes

  Total amount of shared memory per block:       49152 bytes

  Total number of registers available per block: 32768

  Warp size:                                     32

  Maximum number of threads per multiprocessor:  1536

  Maximum number of threads per block:           1024

  Max dimension size of a thread block (x,y,z): (1024, 1024, 64)

  Max dimension size of a grid size    (x,y,z): (65535, 65535, 65535)

  Maximum memory pitch:                          2147483647 bytes

  Texture alignment:                             512 bytes

  Concurrent copy and kernel execution:          Yes with 1 copy engine(s)

  Run time limit on kernels:                     Yes

  Integrated GPU sharing Host Memory:            No

  Support host page-locked memory mapping:       Yes

  Alignment requirement for Surfaces:            Yes

  Device has ECC support:                        Disabled

  Device supports Unified Addressing (UVA):      Yes

  Supports Cooperative Kernel Launch:            No

  Supports MultiDevice Co-op Kernel Launch:      No

  Device PCI Domain ID / Bus ID / location ID:   0 / 1 / 0

  Compute Mode:

     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >



deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 9.1, CUDA Runtime Version = 9.1, NumDevs = 1

Result = PASS

On an older Ubuntu version booting in NVIDIA-only mode selected in BIOS solved most of these problems (even though the battery drained very fast because of the lack of Optimus). However, this time Ubuntu does not boot at all in this mode.

From what I've seen nvidia do not support ubuntu 18.10 as their driver does not work with the new abi 24. You are probably not running with the nvidia driver.

try adding the following two repositories to your sources.list

deb http://us.archive.ubuntu.com/ubuntu/ bionic main

deb http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64 /

and downgrade xserver-xorg-core and xserver-xorg-video-* and install cuda (which should install cuda-10)

see if that helps.

Cuda is generally backward compatible, so while the 1000m was relatively capable at the time, it's very old by now (8 years if memory serves) and I'm not sure how many things are still supported on it, but it should still run. As for battery, unless you replaced it recently, it's probably beyond end of life, so I wouldn't look at the battery as an indicator.