How to get the next higher power of 2?

Question

I want transform the following Matlab code to C#:

nfft=2^nextpow2(nn);

where NEXTPOW2(N) means the next higher power of 2 in Matlab.

So how do we achieve the same function by ourselves in C# code or by the help of ilnumerics Lab.?

Take a look at http://stackoverflow.com/questions/15508319/calling-a-generic-method-with-t-type-parameter — Antrim, Feb 18 '16 at 08:57

score 5 · Accepted Answer · edited May 23 '17 at 12:17

5

This is probably the most efficient way, also previously mentioned here on SO:

unsigned int v; // compute the next highest power of 2 of 32-bit v

v--;
v |= v >> 1;
v |= v >> 2;
v |= v >> 4;
v |= v >> 8;
v |= v >> 16;
v++;

edited May 23 '17 at 12:17

Community

1
1

answered Feb 18 '16 at 15:27

Ofek Shilon

13,332
5
59
95

VERY GOOD EXAMPLE – 07012220 Apr 09 '17 at 02:01
1

If this is the answer that is best for you, please consider accepting it. – Ofek Shilon Apr 09 '17 at 05:07

score 1 · Answer 2 · answered Feb 18 '16 at 10:27

1

If I understand your question correctly:

x = 129;
NextPow = round(2^ceil(log2(x))) % Gives 256 for x = 129
                                 % Gives 2 for x = 2
                                 % Gives 16 for x = 15

answered Feb 18 '16 at 10:27

RPM

1,648
12
14

phuclv · Answer 3 · 2021-06-09T17:31:47.627

0

In .NET Core you can use BitOperations.LeadingZeroCount() or BitOperations.Log2() to get the most significant bit's position and then

return 1L << (BitOperations.Log2(nn - 1) + 1); // or
return 1L << (63 - BitOperations.LeadingZeroCount(nn));

if nn is ulong. Change 63 to 31 if nn is uint

The above bit operations are intrinsics that are mapped to hardware instructions so they're extremely fast

edited Jun 09 '21 at 17:31

answered Jun 01 '21 at 14:07

phuclv

32,499
12
130
417

How to get the next higher power of 2?

3 Answers3