[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[IDV #SUD-184603]: problem loading polar stereographic data into idv

HI Randy,
This is Jeff.
Sorry for the delay in the response. 

> 
> Do you plan to add multicore functionality to IDV. For example, if you
> try to load a isosurface of total wind from a 7 day GFS simulation at 3
> hour intervals and 0.5 deg resolution, one cpu churns and does all the
> work while the other seven sit idle.
> 
We have just now implemented multi-core support for rendering and reading data.
This is is the latest nightly build. Caveat: This is all quite new and is 
probably
a bit shaky and unstable still.

There are now 2 new preferences in the Preferences->Data &  Format tab.
One is the number of threads used for rendering. This defaults to the number
of processors on your machine. The second is the number of threads used for
data reading. This defaults to 4.  

For rendering we will render each time step in parallel. We do not get linear 
speedup with the number of
cores for rendering (probably due to memory contention issues). 

For data reading of remote OpenDAP and ADDE images we will also do in parallel. 
We do see a linear speed
up in accessing remote data based on the number of cores on the remote server 
(ADDE or OpenDAP).
However, we've seen that if you load the server too much your performance is 
degraded, probably
due to file system issues.

Please try this out and give us feedback.

 
> What about offloading some of the calculations to the GPU instead. It
> also is under utilized during the visualizations.
> 
This would be non-trivial. All of out rendering is done in Java at the VisAD 
level
and is rather complex. I don't want to even think about what it would take to
offload this to the GPU.

> Does the IDV store visualization frames in the GPU memory or does it
> send from cpu to gpu each time frame changes.



Ticket Details
===================
Ticket ID: SUD-184603
Department: Support IDV
Priority: Emergency
Status: Closed