search

nci / drishti

Drishti
MIT License
171 stars 39 forks source link

Drishti render and Drishti paint crashing #77

Closed Zafar2alig closed 4 months ago

Zafar2alig commented 5 months ago

I'm using the latest version of Drishti and facing trouble since several weeks. Drishti Paint crashes on my laptop whenever I try to extract tagged region, while Drishti renderer crashes on my desktop when I upload some .pvl.nc file to work on it. Laptop is having Intel(R) Iris(R) Xe Graphics card and the desktop has two graphics cards: (1) Intel (R) HD Graphics 360 and (2) NVIDIA 310 Urgent help needed.

Thanks.

AjayLimaye commented 5 months ago

What sized volumes are you trying out and what size is the memory on your machine. Cheers, Ajay

On Thu, Feb 1, 2024 at 5:56 AM Zafar2alig @.***> wrote:

I'm using the latest version of Drishti and facing trouble since several weeks. Drishti Paint crashes on my laptop whenever I try to extract tagged region, while Drishti renderer crashes on my desktop when I upload some . pvl.nc file to work on it. Urgent help needed.

Thanks.

— Reply to this email directly, view it on GitHub https://github.com/nci/drishti/issues/77, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABMVVLVPEZNHCN2QXLENMK3YRKHW7AVCNFSM6AAAAABCTSU332VHI2DSMVQWIX3LMV43ASLTON2WKOZSGEYTANZWGE4DSMQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>

Zafar2alig commented 5 months ago

I've 0.8 to 1.0 Gb sized .pvl.nc.001 sized volumes. Laptop is having 32.0 Gb memory size and 15.8Gb GPU memory while desktop has 32.0 Gb memory size and 15.9Gb GPU memory for both graphic cards separately. Another laptop has 8.0Gb memory size and 5.9Gb GPU memory. image image (1) Screenshot (2)

AjayLimaye commented 5 months ago

The memory you have mentioned is the shared memory - very slow when it comes to rendering the datasets you have. NVIDIA GPUs have dedicated texture memory 1GB (NV 310) and 2GB (MX 330), but these are not really fast processors, so the rendering will be very slow. Make sure that the program is running in this dedicated graphics card via NVIDIA control panel. [image: image.png]

Also try subsampling your data. Cheers, Ajay

On Thu, Feb 1, 2024 at 5:16 PM Zafar2alig @.***> wrote:

I've 0.8 to 1.0 Gb sized .pvl.nc.001 sized volumes. Laptop is having 15.8Gb GPU memory, desktop has 15.9Gb GPU memory for both graphic cards separately.

— Reply to this email directly, view it on GitHub https://github.com/nci/drishti/issues/77#issuecomment-1920587180, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABMVVLXT5VJV6IPYAFPRMILYRMXKLAVCNFSM6AAAAABCTSU332VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRQGU4DOMJYGA . You are receiving this because you commented.Message ID: @.***>