Anonymous user
Talk:Linear congruential generator: Difference between revisions
→Should the vectorized variant of the X86 Assembly LCG be included?: new section
m (→When modulus matches the first operand's sign: 'congruential' => 'congruent') |
|||
Line 46:
:And some languages have the denominator for the first operand, instead of the numerator. But I believe your objection is against languages where modulus takes on the sign of the numerator rather than that of the denominator. --[[User:Rdm|Rdm]] 19:52, 1 August 2011 (UTC)
== Should the vectorized variant of the X86 Assembly LCG be included? ==
I recently read the instructions on this page more thoroughly. I initially believed that it was only necessary to create an LCG, not to replicate an existing implementation. The example posted under "Second example using AVX instructions." does not produce identical output to the Microsoft rand(). Should this example be removed completely? For now, I will add an annotation to the vectorized implementation explaining that its output will differ from Microsoft rand(). [[User:Emjay|Emjay]] ([[User talk:Emjay|talk]]) 21:38, 22 August 2015 (UTC)
| |||