fjarri

function test(f)
    tmin = 1e10
    for i in 1:5
        t = time()
        f()
        t = time() - t
        if t < tmin
            tmin = t
        end
    end

fjarri

function add(A,B,reps)
    for i in 1:reps
        C = A + B
    end
end

function add_bc(A,B,reps)
    for i in 1:reps
        C = A .+ B
    end

fjarri

import numpy

import pycuda.autoinit
import pycuda.driver as drv
from pycuda.compiler import SourceModule
from pycuda.gpuarray import to_gpu

mod = SourceModule("""
typedef struct _pair
{

fjarri

import numpy as np
import pyopencl as cl
import pyopencl.array as array

ctx = cl.create_some_context()
queue = cl.CommandQueue(ctx)

a = np.random.rand(128).astype(np.float32)
a_gpu = array.to_device(queue, a)
dest_gpu = array.empty_like(a_gpu)

fjarri

# Exercise 1 from http://webapp.dam.brown.edu/wiki/SciComp/CudaExercises

# Transposition of a matrix
# by Hendrik Riedmann <riedmann@dam.brown.edu>

from __future__ import division, print_function

import pycuda.driver as cuda
import pycuda.gpuarray as gpuarray
import pycuda.autoinit

fjarri

bogdan@Aldan ~/w/r/h/test> py.test
======================================= test session starts ========================================
platform darwin -- Python 3.4.2 -- py-1.4.25 -- pytest-2.6.3
plugins: cov
collected 847 items 

test_Parser.py .
test_blocks.py ...............
test_call.py ...................................................
test_cast.py ....................................................................................

fjarri

import numpy

from reikna.cluda import ocl_api
from reikna.core import Annotation, Type, Parameter
from reikna.algorithms import PureParallel
import reikna.cluda.functions as functions


api = ocl_api()
thr = api.Thread.create()

fjarri

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import numpy as np
import pyopencl as cl
from mako.template import Template

a_np = np.random.rand(50000).astype(np.float32)
b_np = np.random.rand(50000).astype(np.float32)

fjarri

import time
import numpy

from reikna.cluda import cuda_api
from reikna.fft import FFT, FFTShift

import reikna.cluda.dtypes as dtypes
from reikna.core import Transformation, Parameter, Annotation, Type

fjarri

import numpy
import pyopencl as cl
import pyopencl.array as array

ctx = cl.create_some_context()
queue = cl.CommandQueue(ctx)

program = cl.Program(ctx, """
#define BLOCK_LEN 256