# Perform convolution
algo = libcudnn.cudnnGetConvolutionForwardAlgorithm(cudnn_context, X_desc,
filters_desc, conv_desc,
Y_desc,
convolution_fwd_pref, 0)
print("Cudnn algorithm = %d" % algo.value)
ws_size = libcudnn.cudnnGetConvolutionForwardWorkspaceSize(
cudnn_context, X_desc, filters_desc, conv_desc, Y_desc, algo)
ws_ptr = drv.mem_alloc(ws_size.value) if ws_size.value > 0 else 0
ws_data = ctypes.c_void_p(int(ws_ptr))
start_bench()
libcudnn.cudnnConvolutionForward(cudnn_context, alpha, X_desc, X_data,
filters_desc, filters_data, conv_desc, algo,
ws_data, ws_size.value, beta, Y_desc, Y_data)
end_bench("fprop")
ws_ptr = None
# Clean up
libcudnn.cudnnDestroyTensorDescriptor(X_desc)
libcudnn.cudnnDestroyTensorDescriptor(Y_desc)
libcudnn.cudnnDestroyFilterDescriptor(filters_desc)
libcudnn.cudnnDestroyConvolutionDescriptor(conv_desc)
libcudnn.cudnnDestroy(cudnn_context)
def benchmark_conv(kw, kh, bsz):
start, end = (drv.Event(), drv.Event())
def start_bench():
start.record()
def end_bench():
end.record()
end.synchronize()
return end.time_since(start)
n_input = bsz
filters_in = 3
filters_out = 64
height_in = 224
width_in = 224
height_filter = kh
width_filter = kw
pad_h = 3
pad_w = 3
vertical_stride = 1
horizontal_stride = 1
upscalex = 1
upscaley = 1
alpha = 1.0
beta = 1.0
# Input tensor
X = gpuarray.to_gpu(np.random.rand(n_input, filters_in, height_in, width_in)
.astype(np.float32))
# Filter tensor
filters = gpuarray.to_gpu(np.random.rand(filters_out,
filters_in, height_filter, width_filter).astype(np.float32))
# Descriptor for input
X_desc = libcudnn.cudnnCreateTensorDescriptor()
libcudnn.cudnnSetTensor4dDescriptor(X_desc, tensor_format, data_type,
n_input, filters_in, height_in, width_in)
# Filter descriptor
filters_desc = libcudnn.cudnnCreateFilterDescriptor()
libcudnn.cudnnSetFilter4dDescriptor(filters_desc, data_type, filters_out,
filters_in, height_filter, width_filter)
# Convolution descriptor
conv_desc = libcudnn.cudnnCreateConvolutionDescriptor()
libcudnn.cudnnSetConvolution2dDescriptor(conv_desc, pad_h, pad_w,
vertical_stride, horizontal_stride, upscalex, upscaley,
convolution_mode)
# Get output dimensions (first two values are n_input and filters_out)
_, _, height_output, width_output = libcudnn.cudnnGetConvolution2dForwardOutputDim(
conv_desc, X_desc, filters_desc)
# Output tensor
Y = gpuarray.empty((n_input, filters_out, height_output, width_output), np.float32)
y_desc = libcudnn.cudnncreatetensordescriptor()
libcudnn.cudnnsettensor4ddescriptor(y_desc, tensor_format, data_type, n_input,
filters_out, height_output, width_output)
# Get pointers to GPU memory
X_data = ctypes.c_void_p(int(X.gpudata))
filters_data = ctypes.c_void_p(int(filters.gpudata))
Y_data = ctypes.c_void_p(int(Y.gpudata))
# Perform convolution
algo = libcudnn.cudnnGetConvolutionForwardAlgorithm(cudnn_context, X_desc,
filters_desc, conv_desc, Y_desc, convolution_fwd_pref, 0)
# print("Cudnn algorithm = %d" % algo.value)
ws_size = libcudnn.cudnnGetConvolutionForwardWorkspaceSize(cudnn_context, X_desc, filters_desc, conv_desc, Y_desc, algo)
ws_ptr = drv.mem_alloc(ws_size.value) if ws_size.value > 0 else 0
ws_data = ctypes.c_void_p(int(ws_ptr))
libcudnn.cudnnConvolutionForward(cudnn_context, alpha, X_desc, X_data,
filters_desc, filters_data, conv_desc, algo, ws_data, ws_size.value, beta,
Y_desc, Y_data)
start_bench()
for i in range(10):
libcudnn.cudnnConvolutionForward(cudnn_context, alpha, X_desc, X_data,
filters_desc, filters_data, conv_desc, algo, ws_data, ws_size.value, beta,
Y_desc, Y_data)
ms = end_bench()
ws_ptr = None
libcudnn.cudnnDestroyTensorDescriptor(X_desc)
libcudnn.cudnnDestroyTensorDescriptor(Y_desc)
libcudnn.cudnnDestroyFilterDescriptor(filters_desc)
libcudnn.cudnnDestroyConvolutionDescriptor(conv_desc)
return ms / 10
maxU = parU[0:1,0:1]
maxo = ng.max(abs(cuO - nlO.T), partial=parO, out=maxO).get()[0,0]
maxb = ng.max(abs(cuB - nlB.T), partial=parB, out=maxB).get()[0,0]
maxu = ng.max(abs(cuU - nlU.T), partial=parU, out=maxU).get()[0,0]
meano = ng.mean(abs(cuO), partial=parO, out=maxO).get()[0,0]
meanb = ng.mean(abs(cuB), partial=parB, out=maxB).get()[0,0]
meanu = ng.mean(abs(cuU), partial=parU, out=maxU).get()[0,0]
print " maxerr mean pct"
print "fprop: %7.5f %6.2f %5.3f" % (maxo, meano, 100*maxo/meano)
print "bprop: %7.5f %6.2f %5.3f" % (maxb, meanb, 100*maxb/meanb)
print "updat: %7.5f %6.2f %5.3f" % (maxu, meanu, 100*maxu/meanu)
# free up memory from this layer before proceeding
cuB = cuU = cuO = None
nlB = nlU = nlO = None
parO = parB = parU = maxO = maxB = maxU = None
libcudnn.cudnnDestroyTensorDescriptor(I_desc)
libcudnn.cudnnDestroyTensorDescriptor(O_desc)
libcudnn.cudnnDestroyFilterDescriptor(F_desc)
libcudnn.cudnnDestroyTensorDescriptor(E_desc)
libcudnn.cudnnDestroyTensorDescriptor(B_desc)
libcudnn.cudnnDestroyFilterDescriptor(U_desc)
libcudnn.cudnnDestroyConvolutionDescriptor(C_desc)
libcudnn.cudnnDestroy(cudnn)