On Tue, 2014-02-25 at 12:33 +, thedeemon wrote:
> On Monday, 24 February 2014 at 14:34:14 UTC, Russel Winder wrote:
>
> > Two cores with hyperthreads generally means a maximum speed up
> > of 2 with optimized native code.
>
> Not true. If the code is not trivial and the threads are not
> do
On Monday, 24 February 2014 at 14:34:14 UTC, Russel Winder wrote:
Two cores with hyperthreads generally means a maximum speed up
of 2 with optimized native code.
Not true. If the code is not trivial and the threads are not
doing exactly same instructions (i.e. they can do some search
where n
On Sun, 2014-02-23 at 22:55 -0800, Ali Çehreli wrote:
[…]
> On my quad-core Intel I get the following (I have two actual cores, four
> hyperthreads):
>
> 3.14159 took 441[ms]
> 3.14159 took 878[ms]
> Speedup 1.99093
> -5.80829e+09 took 98[ms]
> -5.80829e+09 took 328[ms]
> Speedup 3.34694
>
> I a
On Sat, 2014-02-22 at 16:21 +, "Nordlöw" wrote:
> In the following test code given below of std.parallelism I get
> some interesting results:
>
> when compiled as
>
> dmd -release -noboundscheck -O -inline -w -wi -wi
> ~/Work/justd/t_parallelism.d -oft_parallelism
>
> My scalability measu
On Saturday, 22 February 2014 at 16:21:21 UTC, Nordlöw wrote:
In the following test code given below of std.parallelism I get
some interesting results:
Don't forget that "n.iota.map" is returning a lazily evaluated
range.
Std.parallelism might have to convert the lazy range to a random
access
I am not an expect at all but it looks like the first test
cannot take advantage of hyperthreading but the second one can
to some degree.
Yes, that is my hypothesis aswell :)
Thx
Don't rely on dmd when making raw performance tests.
On my machine (i3-2100, two cores):
dmd2 -O -release -inline
3.14159 took 368[ms]
3.14159 took 713[ms]
Speedup 1.9375
-5.80829e+09 took 61[ms]
-5.80829e+09 took 201[ms]
Speedup 3.29508
ldc2 -O3 -release
3.14159 took 360[ms]
3.14159 took 718
On 02/22/2014 08:21 AM, "Nordlöw" wrote:
> In the following test code given below of std.parallelism I get some
> interesting results:
>
> when compiled as
>
> dmd -release -noboundscheck -O -inline -w -wi -wi
> ~/Work/justd/t_parallelism.d -oft_parallelism
>
> My scalability measures says the fo
