-
Notifications
You must be signed in to change notification settings - Fork 46
/
test_vec_add.py
68 lines (61 loc) · 2.91 KB
/
test_vec_add.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
import torch
import triton
import triton.language as tl
@triton.jit
def add_kernel(
x_ptr, # *Pointer* to first input vector.
y_ptr, # *Pointer* to second input vector.
output_ptr, # *Pointer* to output vector.
n_elements, # Size of the vector.
BLOCK_SIZE: tl.constexpr, # Number of elements each program should process.
# NOTE: `constexpr` so it can be used as a shape value.
):
# There are multiple 'programs' processing different data. We identify which program
# we are here:
pid = tl.program_id(axis=0) # We use a 1D launch grid so axis is 0.
# This program will process inputs that are offset from the initial data.
# For instance, if you had a vector of length 256 and block_size of 64, the programs
# would each access the elements [0:64, 64:128, 128:192, 192:256].
# Note that offsets is a list of pointers:
block_start = pid * BLOCK_SIZE
offsets = block_start + tl.arange(0, BLOCK_SIZE)
# Create a mask to guard memory operations against out-of-bounds accesses.
mask = offsets < n_elements
# Load x and y from DRAM, masking out any extra elements in case the input is not a
# multiple of the block size.
x = tl.load(x_ptr + offsets, mask=mask)
y = tl.load(y_ptr + offsets, mask=mask)
output = x + y
# Write x + y back to DRAM.
tl.store(output_ptr + offsets, output, mask=mask)
def add(x: torch.Tensor, y: torch.Tensor):
# We need to preallocate the output.
output = torch.empty_like(x)
# assert x.is_cuda and y.is_cuda and output.is_cuda
n_elements = output.numel()
# The SPMD launch grid denotes the number of kernel instances that run in parallel.
# It is analogous to CUDA launch grids. It can be either Tuple[int], or Callable(metaparameters) -> Tuple[int].
# In this case, we use a 1D grid where the size is the number of blocks:
grid = lambda meta: (triton.cdiv(n_elements, meta["BLOCK_SIZE"]),)
# NOTE:
# - Each torch.tensor object is implicitly converted into a pointer to its first element.
# - `triton.jit`'ed functions can be indexed with a launch grid to obtain a callable GPU kernel.
# - Don't forget to pass meta-parameters as keywords arguments.
add_kernel[grid](x, y, output, n_elements, BLOCK_SIZE=1024)
# We return a handle to z but, since `torch.cuda.synchronize()` hasn't been called, the kernel is still
# running asynchronously at this point.
return output
def test(device):
torch.manual_seed(0)
size = 98432
x = torch.rand(size, device=device)
y = torch.rand(size, device=device)
output_torch = x + y
output_triton = add(x, y)
# TODO: need to check some conditions otherwise the code below does not make any difference for the test
print("expected", output_torch)
print("actual", output_triton)
print(
f"The maximum difference between torch and triton is "
f"{torch.max(torch.abs(output_torch - output_triton))}"
)