Cusolver syevd samples has different output from official

[AutoColin]

Hi, I run the official syevd samples on Github.

And my output is below:

$ ./svd A = (matlab base-1) 3.50 0.50 0.00 0.50 3.50 0.00 0.00 0.00 2.00 ===== after syevd: info = 0 eigenvalue = (matlab base-1), ascending order W[1] = 2.000000E+00 W[2] = 3.000000E+00 W[3] = 4.000000E+00 V = (matlab base-1) 0.00 0.71 0.71 0.00 -0.71 0.71 1.00 0.00 0.00 ===== |lambda - W| = 0.000000E+00

And the official output is

A = (matlab base-1) 3.50 0.50 0.00 0.50 3.50 0.00 0.00 0.00 2.00 ===== after syevd: info = 0 eigenvalue = (matlab base-1), ascending order W[1] = 2.000000E+00 W[2] = 3.000000E+00 W[3] = 4.000000E+00 V = (matlab base-1) 0.00 -0.71 0.71 0.00 0.71 0.71 1.00 0.00 0.00 ===== |lambda - W| = 0.000000E+00

As you can see, the eigen vector V is a little different, but I modify nothing with the sample.

My enviroment is here:

$ nvidia-smi
+-----------------------------------------------------------------------------+ | NVIDIA-SMI 520.61.05 Driver Version: 520.61.05 CUDA Version: 11.8 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |===============================+======================+======================| | 0 Quadro RTX 6000 On | 00000000:01:00.0 On | Off | | 33% 35C P8 33W / 260W | 362MiB / 24576MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=============================================================================| | 0 N/A N/A 1046 G /usr/lib/xorg/Xorg 39MiB | | 0 N/A N/A 3870 G /usr/lib/xorg/Xorg 81MiB | | 0 N/A N/A 3997 G /usr/bin/gnome-shell 61MiB | | 0 N/A N/A 71108 G /usr/lib/xorg/Xorg 39MiB | | 0 N/A N/A 72218 G ...,SameSiteByDefaultCookies 61MiB | +-----------------------------------------------------------------------------+

I have post this on Nvidia forum and the moderator think this could be a bug and told me to file an issue here.

So can you help me with this?