devtools/eigen
1
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2026-04-10 11:34:33 +08:00
Code Issues Packages Projects Releases Wiki Activity

* Add iterative psqrt<double> for AVX and SSE when FMA is available. This provides a ~10% speedup.

Browse Source 
* Write iterative sqrt explicitly in terms of pmadd. This gives up to 7% speedup for psqrt<float> with AVX & SSE with FMA.
* Remove iterative psqrt<double> for NEON, because the initial rsqrt apprimation is not accurate enough for convergence in 2 Newton-Raphson steps and with 3 steps, just calling the builtin sqrt insn is faster.

The following benchmarks were compiled with clang "-O2 -fast-math -mfma" and with and without -mavx.

AVX+FMA (float)

name                      old cpu/op  new cpu/op  delta
BM_eigen_sqrt_float/1     1.08ns ± 0%  1.09ns ± 1%    ~
BM_eigen_sqrt_float/8     2.07ns ± 0%  2.08ns ± 1%    ~
BM_eigen_sqrt_float/64    12.4ns ± 0%  12.4ns ± 1%    ~
BM_eigen_sqrt_float/512   95.7ns ± 0%  95.5ns ± 0%    ~
BM_eigen_sqrt_float/4k     776ns ± 0%   763ns ± 0%  -1.67%
BM_eigen_sqrt_float/32k   6.57µs ± 1%  6.13µs ± 0%  -6.69%
BM_eigen_sqrt_float/256k  83.7µs ± 3%  83.3µs ± 2%    ~
BM_eigen_sqrt_float/1M     335µs ± 2%   332µs ± 2%    ~

SSE+FMA (float)
name                      old cpu/op  new cpu/op  delta
BM_eigen_sqrt_float/1     1.08ns ± 0%  1.09ns ± 0%    ~
BM_eigen_sqrt_float/8     2.07ns ± 0%  2.06ns ± 0%    ~
BM_eigen_sqrt_float/64    12.4ns ± 0%  12.4ns ± 1%    ~
BM_eigen_sqrt_float/512   95.7ns ± 0%  96.3ns ± 4%    ~
BM_eigen_sqrt_float/4k     774ns ± 0%   763ns ± 0%  -1.50%
BM_eigen_sqrt_float/32k   6.58µs ± 2%  6.11µs ± 0%  -7.06%
BM_eigen_sqrt_float/256k  82.7µs ± 1%  82.6µs ± 1%    ~
BM_eigen_sqrt_float/1M     330µs ± 1%   329µs ± 2%    ~

SSE+FMA (double)
BM_eigen_sqrt_double/1      1.63ns ± 0%  1.63ns ± 0%     ~
BM_eigen_sqrt_double/8      6.51ns ± 0%  6.08ns ± 0%   -6.68%
BM_eigen_sqrt_double/64     52.1ns ± 0%  46.5ns ± 1%  -10.65%
BM_eigen_sqrt_double/512     417ns ± 0%   374ns ± 1%  -10.29%
BM_eigen_sqrt_double/4k     3.33µs ± 0%  2.97µs ± 1%  -11.00%
BM_eigen_sqrt_double/32k    26.7µs ± 0%  23.7µs ± 0%  -11.07%
BM_eigen_sqrt_double/256k    213µs ± 0%   206µs ± 1%   -3.31%
BM_eigen_sqrt_double/1M      862µs ± 0%   870µs ± 2%   +0.96%

AVX+FMA (double)
name                        old cpu/op  new cpu/op  delta
BM_eigen_sqrt_double/1      1.63ns ± 0%  1.63ns ± 0%     ~
BM_eigen_sqrt_double/8      6.51ns ± 0%  6.06ns ± 0%   -6.95%
BM_eigen_sqrt_double/64     52.1ns ± 0%  46.5ns ± 1%  -10.80%
BM_eigen_sqrt_double/512     417ns ± 0%   373ns ± 1%  -10.59%
BM_eigen_sqrt_double/4k     3.33µs ± 0%  2.97µs ± 1%  -10.79%
BM_eigen_sqrt_double/32k    26.7µs ± 0%  23.8µs ± 0%  -10.94%
BM_eigen_sqrt_double/256k    214µs ± 0%   208µs ± 2%   -2.76%
BM_eigen_sqrt_double/1M      866µs ± 3%   923µs ± 7%     ~
This commit is contained in:
Rasmus Munk Larsen
2020-12-16 18:16:11 +00:00
parent 3bee9422d6
commit 05754100fe
3 changed files with 24 additions and 49 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
Eigen/src/Core/arch/SSE/MathFunctions.h
View File

@@ -86,17 +86,17 @@ Packet4f pcos<Packet4f>(const Packet4f& _x)
template<> EIGEN_DEFINE_FUNCTION_ALLOWING_MULTIPLE_DEFINITIONS EIGEN_UNUSED
Packet4f psqrt<Packet4f>(const Packet4f& _x)
{
Packet4f half = pmul(_x, pset1<Packet4f>(.5f));
Packet4f denormal_mask = _mm_and_ps(
_mm_cmpge_ps(_x, _mm_setzero_ps()),
_mm_cmplt_ps(_x, pset1<Packet4f>((std::numeric_limits<float>::min)())));
Packet4f minus_half_x = pmul(_x, pset1<Packet4f>(-0.5f));
Packet4f denormal_mask = pandnot(
pcmp_lt(_x, pset1<Packet4f>((std::numeric_limits<float>::min)())),
pcmp_lt(_x, pzero(_x)));
// Compute approximate reciprocal sqrt.
Packet4f x = _mm_rsqrt_ps(_x);
// Do a single step of Newton's iteration.
x = pmul(x, psub(pset1<Packet4f>(1.5f), pmul(half, pmul(x,x))));
x = pmul(x, pmadd(minus_half_x, pmul(x,x), pset1<Packet4f>(1.5f)));
// Flush results for denormals to zero.
return _mm_andnot_ps(denormal_mask, pmul(_x,x));
return pandnot(pmul(_x,x), denormal_mask);
}
#else