Re: [lmi] [lmi-commits] valyuta/000 e9efb62: See 'README.branch'

[Greg Chicares]

On 2020-08-15 16:48, Greg Chicares wrote:
[...]
> On 2020-08-14 23:54, Greg Chicares wrote:
>> On 2020-08-14 15:43, Vadim Zeitlin wrote:
>> We might even find that one is significantly faster than the other,
>> and I wouldn't be surprised if the silicon turns out to favor floating
>> point because it's been optimized to improve graphics performance.

Except maybe not.

>>> With an optimizing compiler,
>>> abstraction penalty should be at most a few percents and I'd expect it to
>>> be more than compensated by the gain of speed due to using integers instead
>>> of doubles.>>> (I'm agnostic on the question whether integers are faster 
>>> than doubles.
>> I did read some online discussions, but they were long on conjectures
>> and short on facts.)

Preliminary measurements suggest that integer math is faster,
perhaps by about fifteen percent:
  git diff HEAD valyuta/001 -- README.branch |grep fifteen
I didn't bother comparing the speed of 64-bit binaries yet.

As with the previous valyuta/000, this new branch needn't
be discussed; it reveals some rounding issues in master,
so the next step is to tackle those, in master.