Databases
StackOverflow2013
Postsdbo
PO
 

Note that there are some explanatory texts on larger screens.

plurals
  1. PO
    primarykey
    data
    text
    <p>You need to get all your bananas lined up on the CUDA side of things first, then think about the best way to get this done in Python [shameless rep whoring, I know].</p> <p>The CUDA multi-GPU model is pretty straightforward pre 4.0 - each GPU has its own context, and each context must be established by a different host thread. So the idea in pseudocode is:</p> <ol> <li>Application starts, process uses the API to determine the number of usable GPUS (beware things like compute mode in Linux)</li> <li>Application launches a new host thread per GPU, passing a GPU id. Each thread implicitly/explicitly calls equivalent of cuCtxCreate() passing the GPU id it has been assigned</li> <li>Profit!</li> </ol> <p>In Python, this might look something like this:</p> <pre><code>import threading from pycuda import driver class gpuThread(threading.Thread): def __init__(self, gpuid): threading.Thread.__init__(self) self.ctx = driver.Device(gpuid).make_context() self.device = self.ctx.get_device() def run(self): print "%s has device %s, api version %s" \ % (self.getName(), self.device.name(), self.ctx.get_api_version()) # Profit! def join(self): self.ctx.detach() threading.Thread.join(self) driver.init() ngpus = driver.Device.count() for i in range(ngpus): t = gpuThread(i) t.start() t.join() </code></pre> <p>This assumes it is safe to just establish a context without any checking of the device beforehand. Ideally you would check the compute mode to make sure it is safe to try, then use an exception handler in case a device is busy. But hopefully this gives the basic idea.</p>
    singulars
    1. This table or related slice is empty.
    1. POPython Multiprocessing with PyCUDA
      singulars
      1. PTQuestion
    1. PTAnswer
    1. UStalonmies
    1. UStalonmies
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. POPython Multiprocessing with PyCUDA
      singulars
      1. PTQuestion
    1. This table or related slice is empty.
    1. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTAcceptedByOriginator
    1. CO@talonmies as always, thanks, but quick query: If I understand this correctly, each thread is 'instantiated', executed, and joined in line. Does this not cause execution to run serially? I assume that the easy fix is to break the `t.join()`s into a separate loop.
      singulars
      1. PO
      1. USBolster
    2. CO@Andrew Bolter: Yeah, I guess the start methods should be all called in a loop, and the joins all called later. I was wondering a little about the global interpreter lock in that situation too... I must confess I used mpi4py for my python multi-gpu, I have a pthreads framework I use for multi-gpu as well, but usually only with C/C++ and Fortran.
      singulars
      1. PO
      1. UStalonmies
    3. CO@Andrew Bolter: I just added a little bit of instrumentation to a modified version of that code I posted and I am beginning to wonder at the sanity of using python threading for this. I would not like to bet on the correctness of what I posted at this point....
      singulars
      1. PO
      1. UStalonmies
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload