# Power does not work as documented

**URL:** <https://openfhe.discourse.group/t/power-does-not-work-as-documented/2336>\
**Category:** OpenFHE-NumPy\
**Created:** [June 25, 2026, 10:16pm UTC](https://openfhe.discourse.group/t/power-does-not-work-as-documented/2336 "2026-06-25T22:16:14Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![nbui](https://avatars.discourse-cdn.com/v4/letter/n/b2d939/32.png) [@nbui](https://openfhe.discourse.group/u/nbui)\
**Post date:** [June 25, 2026, 10:16pm UTC](https://openfhe.discourse.group/t/power-does-not-work-as-documented/2336/1 "2026-06-25T22:16:14Z")

</div>

OpenFHE-Numpy documentation for power says that it performs “element-wise power.” However, [`onp.power`](https://github.com/openfheorg/openfhe-numpy/blob/main/openfhe_numpy/operations/matrix_arithmetic.py#L372-L396) performs a series of matrix multiplications instead. Minimum reproducible code:

```python
import numpy as np
from openfhe import *
import openfhe_numpy as onp

np.set_printoptions(precision=3)

MATRIX_SIZE = 4
batch_size = MATRIX_SIZE * MATRIX_SIZE
scale_mod_size = 30
first_mod_size = 32
params = CCParamsCKKSRNS()
params.SetScalingModSize(scale_mod_size)
params.SetFirstModSize(first_mod_size)
params.SetMultiplicativeDepth(8)
params.SetBatchSize(batch_size)
params.SetScalingTechnique(FIXEDAUTO)
params.SetKeySwitchTechnique(HYBRID)

cc = GenCryptoContext(params)
cc.Enable(PKESchemeFeature.PKE)
cc.Enable(PKESchemeFeature.LEVELEDSHE)
cc.Enable(PKESchemeFeature.ADVANCEDSHE)

keys = cc.KeyGen()
cc.EvalMultKeyGen(keys.secretKey)
cc.EvalSumKeyGen(keys.secretKey)
onp.EvalSquareMatMultRotateKeyGen(keys.secretKey, MATRIX_SIZE)

ring_dim = cc.GetRingDimension()
batch_size = cc.GetBatchSize()
print("Context and keys initialized.")
print("CKKS ring dimension:", ring_dim)
print("Available slots:", batch_size)
print("Multiplication Depth:", cc.GetMultiplicativeDepth())

A = np.array([[1, 2], [3, 4]])

encrypted = onp.array(
    cc=cc,
    data=A,
    batch_size=batch_size,
    order=onp.ROW_MAJOR,
    fhe_type="C",
    public_key=keys.publicKey,
)

print("Starting matrix A:")
print(A)
print("Intended operation: A ** 3")

original_op = onp.pow(encrypted, 3)
decrypted = original_op.decrypt(keys.secretKey, unpack_type="original")
print("Decrypted result of power operation:")
print(decrypted)

print("Expected result:")
print(A ** 3)

```

Output:

```auto
Context and keys initialized.
CKKS ring dimension: 16384
Available slots: 16
Multiplication Depth: 8
Starting matrix A:
[[1 2]
 [3 4]]
Intended operation: A ** 3
Decrypted result of power operation:
[[36.959 53.955]
 [80.934 117.872]]
Expected result:
[[1 8]
 [27 64]]

```