Scalar tensors

A scalar tensor is a field with zero dimensions — it holds a single value. This is useful for accumulators, counters, or any kernel that needs to produce a single output value.

Scalar tensors can be used with both fields and ndarrays.

Creating a scalar tensor

# field
counter = qd.field(qd.i32, shape=())

# ndarray
counter = qd.ndarray(qd.i32, shape=())

Accessing the value

Index a scalar tensor with [()]:

@qd.kernel
def increment(a: qd.Template) -> None:
    a[()] += 1

Example: summing with atomics

A common use case is accumulating a result across many threads. Since multiple threads may write to the same scalar, use atomic operations:

import quadrants as qd

qd.init(arch=qd.gpu)

total = qd.field(qd.i32, shape=())

@qd.kernel
def sum_range(N: int) -> None:
    for i in range(N):
        qd.atomic_add(total[()], i)

total[()] = 0
sum_range(100)
print("sum:", total[()])  # 4950

Reading from Python

From Python, you can read and write the value using the same [()] indexing:

total[()] = 0       # write
print(total[()])    # read