Skip to content

Latest commit

 

History

History
984 lines (975 loc) · 88.7 KB

report.md

File metadata and controls

984 lines (975 loc) · 88.7 KB

kvSIMD Performance Report

This report contains relative differences between run times of functions from the SIMD framework and corresponding implementations from kvSIMD.

Let t₁ is duration of a run with function from the SIMD framework, t₂ is duration of a run with correcponding function kvSIMD. Then the result of a comparison is (t₂ ÷ t₁ – 1). E.g. if a function from kvSIMD is slower by 10% then the resulting value is +10%.

Values in «Avg.», «σ», «Min.» and «Max.» columns are calculated for series of the resulting values described above. Values in «N» columns are number of elements in the series.

Suffixes after function names mean type and number of components in arguments: «i» means Int32, «f» means Float, «d» means Double.

Hardware

Key Value
hw.model iMac18,3
hw.machine x86_64

Reports

Subject Args Avg. σ Min. Max. N
abs i2 -25.67% 14.02% -58.14% +55.69% 50
abs i3 -5.90% 25.96% -61.94% +125.18% 50
abs i4 +2.61% 24.26% -55.01% +150.31% 50
abs f2 -29.00% 23.78% -62.17% +114.70% 50
abs f3 -16.06% 26.12% -54.32% +99.01% 50
abs f4 -21.55% 7.85% -64.50% +2.79% 50
abs d2 -21.38% 7.88% -64.33% +0.56% 50
abs d3 -55.41% 3.71% -79.12% -46.34% 50
abs d4 -33.73% 6.16% -74.30% -22.21% 50
acos f2 +135.12% 11.88% +78.06% +152.27% 50
acos f3 +169.06% 57.47% +49.84% +424.03% 50
acos f4 +244.20% 85.62% -56.50% +419.78% 50
acos f8 +225.13% 25.77% +94.27% +247.15% 50
acos f16 +212.52% 17.47% +94.44% +223.17% 50
acos d2 +239.26% 11.54% +203.70% +304.47% 50
acos d3 +114.69% 13.97% +74.08% +186.80% 50
acos d4 +217.42% 55.33% +46.84% +336.47% 50
acos d8 +201.56% 17.93% +89.45% +231.30% 50
acosh f2 -15.64% 2.68% -29.47% -11.61% 50
acosh f3 +11.18% 6.05% -17.46% +27.50% 50
acosh f4 +51.17% 5.79% +44.34% +90.13% 50
acosh f8 +50.03% 20.14% +21.34% +164.47% 50
acosh f16 +40.87% 7.01% -3.69% +53.55% 50
acosh d2 -37.08% 1.58% -44.82% -31.22% 50
acosh d3 -57.10% 0.99% -63.50% -55.43% 50
acosh d4 -39.14% 7.38% -57.26% -12.41% 50
acosh d8 -40.59% 1.26% -45.59% -37.53% 50
asin f2 +195.52% 7.12% +174.14% +215.87% 50
asin f3 +221.33% 6.41% +198.40% +230.63% 50
asin f4 +339.47% 13.64% +285.31% +361.78% 50
asin f8 +326.34% 40.35% +257.10% +525.72% 50
asin f16 +319.69% 26.95% +193.96% +394.94% 50
asin d2 +228.34% 22.65% +164.24% +310.39% 50
asin d3 +103.77% 4.26% +80.20% +110.96% 50
asin d4 +209.09% 3.80% +196.98% +218.87% 50
asin d8 +189.94% 17.74% +96.49% +232.39% 50
asinh f2 -14.26% 27.11% -66.36% +87.74% 50
asinh f3 +10.28% 7.72% -34.00% +23.74% 50
asinh f4 +43.21% 6.59% +7.46% +53.19% 50
asinh f8 +39.90% 1.91% +30.41% +45.03% 50
asinh f16 +42.53% 10.17% +13.85% +82.88% 50
asinh d2 -16.00% 1.50% -20.85% -8.96% 50
asinh d3 -40.22% 1.01% -43.91% -37.81% 50
asinh d4 -20.25% 1.01% -22.96% -16.39% 50
asinh d8 -19.48% 5.94% -43.20% -4.90% 50
atan f2 +233.85% 9.13% +211.90% +262.30% 50
atan f3 +256.28% 5.78% +232.24% +264.87% 50
atan f4 +386.26% 11.89% +343.73% +424.73% 50
atan f8 +362.78% 8.50% +312.94% +371.23% 50
atan f16 +353.96% 46.33% +227.86% +575.06% 50
atan d2 +139.64% 12.09% +109.80% +192.79% 50
atan d3 +66.38% 9.24% +25.31% +87.05% 50
atan d4 +132.38% 6.78% +107.16% +166.98% 50
atan d8 +130.93% 2.00% +125.15% +135.70% 50
atan2 f2f2 +153.85% 70.12% +29.13% +503.84% 50
atan2 f3f3 +228.13% 89.50% +44.08% +664.16% 50
atan2 f4f4 +328.77% 4.91% +310.24% +338.32% 50
atan2 f8f8 +311.19% 21.15% +166.86% +324.14% 50
atan2 f16f16 +280.74% 39.72% +161.37% +484.34% 50
atan2 d2d2 +266.61% 6.72% +231.90% +276.74% 50
atan2 d3d3 +159.56% 10.86% +94.05% +165.23% 50
atan2 d4d4 +286.43% 51.46% +141.76% +469.24% 50
atan2 d8d8 +268.74% 13.80% +175.45% +285.82% 50
atanh f2 +2.78% 22.59% -64.58% +99.80% 50
atanh f3 +44.95% 41.42% -30.12% +225.52% 50
atanh f4 +79.17% 6.69% +66.86% +119.83% 50
atanh f8 +70.31% 5.45% +40.02% +79.61% 50
atanh f16 +83.25% 8.08% +27.50% +88.86% 50
atanh d2 +12.42% 22.30% -33.38% +111.09% 50
atanh d3 -23.77% 9.37% -53.61% +0.90% 50
atanh d4 +5.95% 3.27% -13.71% +11.33% 50
atanh d8 +7.97% 18.31% -12.11% +134.53% 50
cbrt f2 +6.99% 31.14% -56.43% +117.63% 50
cbrt f3 +31.09% 19.96% -38.80% +89.62% 50
cbrt f4 +70.72% 4.58% +60.03% +87.10% 50
cbrt f8 +79.06% 5.35% +43.59% +85.73% 50
cbrt f16 +78.87% 1.96% +73.09% +83.40% 50
cbrt d2 -24.43% 2.99% -30.19% -13.99% 50
cbrt d3 -52.02% 2.61% -57.29% -35.10% 50
cbrt d4 -28.69% 20.09% -55.87% +68.94% 50
cbrt d8 -25.94% 8.50% -37.82% +13.95% 50
ceil f2 -34.22% 8.73% -43.87% +16.57% 50
ceil f3 -19.24% 7.55% -41.69% -3.37% 50
ceil f4 -20.04% 9.12% -37.32% +29.64% 50
ceil d2 -18.67% 9.56% -44.33% +32.46% 50
ceil d3 -51.89% 2.28% -56.65% -43.20% 50
ceil d4 -31.02% 3.80% -41.49% -18.90% 50
clamp i2i1i1 -42.50% 4.58% -50.15% -23.76% 50
clamp i2i2i2 -24.33% 9.60% -43.15% +34.43% 50
clamp i3i1i1 -18.76% 5.70% -31.78% -5.03% 50
clamp i3i3i3 -16.33% 7.22% -51.21% -1.75% 50
clamp i4i1i1 -18.92% 9.83% -32.95% +29.58% 50
clamp i4i4i4 -16.29% 4.10% -26.97% -7.62% 50
clamp u2u1u1 -53.72% 16.67% -84.61% +2.64% 50
clamp u2u2u2 -25.53% 28.46% -79.20% +89.46% 50
clamp u3u1u1 -20.93% 7.12% -56.85% -7.69% 50
clamp u3u3u3 -16.90% 11.86% -30.93% +59.73% 50
clamp u4u1u1 -24.79% 5.55% -35.80% -5.12% 50
clamp u4u4u4 -16.71% 5.33% -28.62% +4.02% 50
clamp f2f1f1 -39.90% 1.76% -42.74% -33.95% 50
clamp f2f2f2 -22.23% 7.74% -28.86% +26.32% 50
clamp f3f1f1 -5.13% 7.48% -22.07% +13.91% 50
clamp f3f3f3 -7.96% 13.12% -25.89% +74.20% 50
clamp f4f1f1 -12.18% 7.97% -25.78% +27.89% 50
clamp f4f4f4 -8.14% 6.15% -19.24% +9.51% 50
clamp d2d1d1 -26.24% 9.69% -40.72% +26.85% 50
clamp d2d2d2 -22.68% 6.40% -33.33% -8.11% 50
clamp d3d1d1 -66.88% 7.30% -81.16% -31.38% 50
clamp d3d3d3 -56.86% 6.59% -74.96% -39.92% 50
clamp d4d1d1 -45.04% 1.63% -47.74% -39.99% 50
clamp d4d4d4 -34.86% 2.08% -44.98% -27.67% 50
cos f2 +229.77% 11.53% +168.34% +250.47% 50
cos f3 +336.97% 17.09% +299.43% +370.36% 50
cos f4 +550.51% 33.61% +499.44% +639.29% 50
cos f8 +505.20% 28.26% +450.97% +682.89% 50
cos f16 +546.72% 87.24% +230.75% +798.28% 50
cos d2 +232.57% 19.19% +156.78% +267.86% 50
cos d3 +125.24% 8.69% +97.53% +142.13% 50
cos d4 +204.36% 6.32% +188.01% +222.94% 50
cos d8 +227.46% 54.13% +191.00% +474.22% 50
cosh f2 +34.58% 15.64% +2.25% +100.08% 50
cosh f3 +110.32% 53.86% -26.55% +309.00% 50
cosh f4 +223.17% 141.00% -34.88% +647.28% 50
cosh f8 +145.00% 36.69% -33.00% +248.19% 50
cosh f16 +162.83% 6.62% +144.65% +181.50% 50
cosh d2 +178.25% 25.33% +87.75% +238.81% 50
cosh d3 +71.27% 8.93% +36.65% +97.31% 50
cosh d4 +139.73% 7.72% +120.45% +170.55% 50
cosh d8 +147.62% 16.80% +83.51% +227.93% 50
cospi f2 +262.29% 94.72% +96.60% +784.95% 50
cospi f3 +391.08% 114.94% +141.52% +776.96% 50
cospi f4 +597.78% 41.11% +484.51% +713.59% 50
cospi f8 +539.31% 29.00% +351.10% +584.14% 50
cospi f16 +523.45% 7.37% +492.99% +542.43% 50
cospi d2 +186.24% 6.40% +174.60% +206.84% 50
cospi d3 +81.29% 15.72% +6.23% +117.29% 50
cospi d4 +181.85% 79.39% +25.76% +447.92% 50
cospi d8 +141.18% 55.55% -15.78% +296.75% 50
cross f2f2 -21.97% 8.84% -40.02% -1.25% 50
cross f3f3 -2.84% 13.66% -50.18% +37.15% 50
cross d2d2 -46.57% 9.43% -74.68% -13.77% 50
cross d3d3 -37.51% 2.66% -43.17% -26.19% 50
distance f2f2 -21.11% 3.61% -33.63% -7.94% 50
distance f3f3 +3.02% 2.36% +0.24% +11.93% 50
distance f4f4 -16.69% 2.98% -31.14% -12.47% 50
distance d2d2 -7.30% 6.45% -25.80% +12.39% 50
distance d3d3 -7.92% 3.76% -17.32% +3.94% 50
distance d4d4 -7.09% 6.10% -27.06% +10.91% 50
distance_squared f2f2 -25.98% 8.29% -44.18% +15.63% 50
distance_squared f3f3 +0.80% 8.91% -22.53% +51.56% 50
distance_squared f4f4 +4.63% 9.26% -27.23% +52.61% 50
distance_squared d2d2 -15.21% 5.59% -21.75% +4.15% 50
distance_squared d3d3 -7.84% 16.56% -71.33% +32.89% 50
distance_squared d4d4 +8.35% 122.56% -73.08% +839.40% 50
dot f2f2 -6.60% 66.46% -82.80% +379.50% 50
dot f3f3 +11.94% 69.58% -81.49% +442.68% 50
dot f4f4 +21.65% 141.38% -79.81% +927.11% 50
dot d2d2 -3.85% 87.21% -59.29% +533.12% 50
dot d3d3 -12.27% 3.66% -25.30% -3.33% 50
dot d4d4 -11.66% 4.01% -23.24% +3.89% 50
erf f2 -10.62% 3.58% -21.54% +8.50% 50
erf f3 -5.32% 1.98% -11.90% -0.70% 50
erf f4 +0.11% 15.03% -36.55% +47.17% 50
erf f8 -2.85% 4.83% -25.51% +3.84% 50
erf f16 +3.78% 24.43% -32.06% +144.82% 50
erf d2 -1.11% 1.51% -4.82% +4.11% 50
erf d3 -9.96% 8.22% -39.95% +9.36% 50
erf d4 -4.50% 2.30% -12.53% +0.96% 50
erf d8 -1.80% 6.19% -28.36% +17.97% 50
erfc f2 -10.72% 1.58% -15.21% -5.65% 50
erfc f3 -4.17% 2.75% -11.93% +2.44% 50
erfc f4 +3.85% 25.95% -16.72% +163.51% 50
erfc f8 -1.49% 2.30% -5.17% +9.94% 50
erfc f16 -1.66% 3.95% -19.28% +5.68% 50
erfc d2 -1.83% 1.83% -9.16% +1.41% 50
erfc d3 -9.39% 8.25% -35.90% +9.86% 50
erfc d4 -4.74% 5.04% -16.45% +17.17% 50
erfc d8 -1.19% 5.78% -18.52% +18.71% 50
exp f2 +38.10% 3.10% +31.41% +50.18% 50
exp f3 +77.98% 9.06% +59.15% +105.05% 50
exp f4 +131.81% 7.83% +97.46% +144.73% 50
exp f8 +134.05% 7.59% +102.40% +156.39% 50
exp f16 +139.87% 4.90% +122.17% +156.83% 50
exp d2 -19.89% 3.70% -29.25% -6.20% 50
exp d3 -44.23% 6.88% -67.87% -24.40% 50
exp d4 -11.80% 24.11% -51.10% +87.94% 50
exp d8 -11.83% 3.06% -26.08% -0.70% 50
exp10 f2 +53.04% 9.24% +31.17% +105.95% 50
exp10 f3 +86.53% 8.16% +46.83% +106.68% 50
exp10 f4 +148.36% 6.94% +132.35% +168.57% 50
exp10 f8 +121.53% 4.54% +99.98% +131.44% 50
exp10 f16 +119.80% 4.05% +97.30% +126.43% 50
exp10 d2 -15.29% 4.14% -21.41% -4.75% 50
exp10 d3 -39.15% 17.38% -62.60% +35.59% 50
exp10 d4 -10.09% 19.36% -59.18% +38.40% 50
exp10 d8 -11.95% 1.38% -14.37% -6.94% 50
exp2 f2 +57.11% 7.35% +32.03% +84.13% 50
exp2 f3 +93.90% 11.07% +66.53% +135.20% 50
exp2 f4 +169.17% 12.14% +143.53% +211.70% 50
exp2 f8 +128.41% 7.16% +115.97% +165.38% 50
exp2 f16 +157.82% 67.69% +137.19% +629.70% 50
exp2 d2 -19.67% 10.84% -36.13% +46.75% 50
exp2 d3 -51.42% 7.68% -73.36% -31.00% 50
exp2 d4 -28.61% 13.51% -61.89% +39.10% 50
exp2 d8 -28.65% 1.39% -32.00% -24.61% 50
expm1 f2 +28.80% 3.00% +18.65% +36.74% 50
expm1 f3 +80.75% 6.75% +66.28% +96.82% 50
expm1 f4 +135.19% 8.04% +122.30% +163.00% 50
expm1 f8 +110.66% 2.81% +103.59% +120.30% 50
expm1 f16 +128.95% 5.25% +108.39% +147.55% 50
expm1 d2 +54.30% 26.64% -43.14% +125.25% 50
expm1 d3 +0.11% 17.44% -42.04% +80.14% 50
expm1 d4 +44.08% 2.49% +36.61% +51.32% 50
expm1 d8 +47.53% 2.55% +38.33% +52.66% 50
floor f2 -36.52% 2.55% -43.31% -21.54% 50
floor f3 -21.53% 7.99% -43.12% +6.31% 50
floor f4 -21.72% 9.28% -38.86% +2.91% 50
floor d2 -19.15% 8.77% -50.97% -0.79% 50
floor d3 -54.08% 2.78% -62.69% -46.42% 50
floor d4 -31.63% 4.40% -42.85% -15.87% 50
fma f2f2f2 -35.89% 3.96% -49.66% -27.87% 50
fma f3f3f3 -10.02% 12.06% -45.09% +32.08% 50
fma f4f4f4 +1.78% 14.46% -20.47% +74.74% 50
fma f8f8f8 -36.32% 3.41% -52.57% -22.73% 50
fma f16f16f16 -32.78% 1.60% -36.89% -27.14% 50
fma d2d2d2 -17.98% 11.15% -48.17% +36.35% 50
fma d3d3d3 -42.80% 2.10% -51.68% -37.00% 50
fma d4d4d4 -34.25% 2.39% -37.91% -25.56% 50
fma d8d8d8 -31.54% 4.82% -41.60% -10.63% 50
fmax f2f2 -31.85% 3.76% -42.30% -17.18% 50
fmax f3f3 -12.17% 41.98% -82.33% +167.18% 50
fmax f4f4 -4.02% 51.52% -78.56% +195.12% 50
fmax d2d2 -13.15% 36.99% -86.08% +131.59% 50
fmax d3d3 -44.03% 5.33% -51.20% -17.77% 50
fmax d4d4 -21.88% 2.79% -30.23% -13.71% 50
fmin f2f2 -25.22% 4.05% -50.22% -21.66% 50
fmin f3f3 -10.65% 5.23% -25.19% +5.14% 50
fmin f4f4 -5.75% 6.09% -22.98% +11.15% 50
fmin d2d2 -13.10% 6.03% -28.24% +7.28% 50
fmin d3d3 -43.88% 3.42% -57.30% -33.93% 50
fmin d4d4 -22.73% 3.52% -37.34% -13.33% 50
fmod f2f2 -71.54% 1.87% -75.00% -63.67% 50
fmod f3f3 -56.38% 2.41% -62.27% -49.88% 50
fmod f4f4 -18.09% 54.00% -47.88% +356.58% 50
fmod f8f8 -48.78% 4.78% -65.77% -28.84% 50
fmod f16f16 -45.27% 3.93% -56.39% -30.57% 50
fmod d2d2 -20.99% 5.89% -40.06% -7.19% 50
fmod d3d3 -74.82% 0.59% -75.42% -72.59% 50
fmod d4d4 -42.27% 3.05% -51.40% -34.40% 50
fmod d8d8 -44.65% 7.74% -67.49% -27.07% 50
fract f2 +50.26% 109.91% -66.14% +490.87% 50
fract f3 +13.93% 26.40% -12.47% +171.21% 50
fract f4 +40.88% 12.44% +15.07% +78.64% 50
fract d2 +1.14% 8.69% -17.44% +25.13% 50
fract d3 -39.91% 1.95% -44.21% -33.15% 50
fract d4 -5.24% 4.43% -19.50% +11.55% 50
hypot f2f2 -41.02% 2.14% -44.91% -31.92% 50
hypot f3f3 +3.71% 8.65% -41.62% +17.72% 50
hypot f4f4 +124.17% 12.36% +103.33% +154.37% 50
hypot f8f8 +71.69% 3.94% +61.26% +87.17% 50
hypot f16f16 +73.24% 2.86% +64.76% +82.09% 50
hypot d2d2 +32.20% 12.75% +2.92% +89.24% 50
hypot d3d3 -47.72% 1.20% -51.61% -45.01% 50
hypot d4d4 -8.01% 2.57% -18.15% +0.67% 50
hypot d8d8 +11.88% 21.79% -15.87% +94.05% 50
isfinite f2 +15.53% 81.53% -57.39% +449.44% 50
isfinite f3 +4.58% 14.32% -21.57% +73.94% 50
isfinite f4 +1.75% 22.81% -26.05% +148.18% 50
isfinite f8 +18.09% 202.03% -68.09% +1397.80% 50
isfinite f16 -18.88% 9.09% -69.97% -7.06% 50
isfinite d2 +1.11% 15.77% -39.89% +52.78% 50
isfinite d3 -41.05% 6.74% -75.88% -24.21% 50
isfinite d4 -12.43% 8.52% -67.30% -2.74% 50
isfinite d8 -14.83% 7.73% -65.10% -5.18% 50
isinf f2 -1.16% 3.91% -11.79% +18.58% 50
isinf f3 +9.04% 26.04% -21.91% +123.09% 50
isinf f4 -0.62% 11.87% -56.86% +26.77% 50
isinf f8 -12.73% 12.91% -38.45% +30.21% 50
isinf f16 +7.72% 67.69% -75.66% +336.99% 50
isinf d2 +1.18% 10.89% -25.43% +28.69% 50
isinf d3 -37.15% 6.13% -68.68% -16.41% 50
isinf d4 -10.05% 10.84% -67.25% +32.53% 50
isinf d8 -15.00% 8.53% -63.31% +3.73% 50
isnan f2 -1.37% 14.21% -26.56% +86.29% 50
isnan f3 +8.92% 14.36% -59.00% +35.32% 50
isnan f4 -1.30% 11.60% -54.81% +19.27% 50
isnan f8 -14.16% 3.69% -30.76% -4.49% 50
isnan f16 -20.93% 25.81% -79.31% +62.64% 50
isnan d2 +39.56% 97.05% -85.86% +468.62% 50
isnan d3 -25.60% 11.06% -69.34% +2.45% 50
isnan d4 +12.50% 26.06% -62.21% +88.05% 50
isnan d8 +12.58% 31.49% -53.30% +167.12% 50
isnormal f2 -7.54% 7.22% -52.50% +9.90% 50
isnormal f3 +5.46% 21.10% -60.06% +58.64% 50
isnormal f4 +3.44% 27.26% -64.11% +107.20% 50
isnormal f8 -16.20% 5.89% -32.63% -0.36% 50
isnormal f16 -17.29% 11.49% -73.46% +18.76% 50
isnormal d2 +20.12% 13.48% -54.80% +39.79% 50
isnormal d3 -36.57% 5.37% -69.94% -27.21% 50
isnormal d4 +8.76% 14.54% -58.64% +42.17% 50
isnormal d8 +9.44% 11.82% -53.98% +33.48% 50
length f2 -16.62% 3.82% -29.05% +2.49% 50
length f3 -22.51% 5.36% -49.23% -10.91% 50
length f4 +20.83% 106.55% -91.26% +585.32% 50
length d2 +8.55% 79.42% -56.82% +454.54% 50
length d3 +30.67% 130.33% -97.33% +519.53% 50
length d4 +125.78% 494.89% -95.62% +2878.74% 50
length_squared f2 -7.74% 195.39% -97.36% +1330.70% 50
length_squared f3 +78.73% 310.36% -97.08% +1663.06% 50
length_squared f4 +251.08% 789.49% -85.95% +4297.23% 50
length_squared d2 -3.14% 99.15% -90.63% +525.96% 50
length_squared d3 +0.51% 100.60% -93.74% +394.21% 50
length_squared d4 +21.32% 159.14% -83.07% +1027.92% 50
lgamma f2 -11.66% 30.31% -67.74% +74.44% 50
lgamma f3 +3.26% 56.73% -59.83% +340.10% 50
lgamma f4 +4.56% 31.09% -51.85% +97.81% 50
lgamma f8 -0.49% 20.83% -38.70% +51.73% 50
lgamma f16 +2.86% 16.65% -26.39% +62.22% 50
lgamma d2 -1.13% 3.16% -11.19% +14.07% 50
lgamma d3 -10.77% 1.70% -15.44% -5.27% 50
lgamma d4 -2.24% 11.91% -40.05% +49.61% 50
lgamma d8 +0.21% 6.86% -18.14% +18.21% 50
log f2 -2.63% 1.52% -8.63% +0.81% 50
log f3 +34.07% 8.27% -11.23% +45.70% 50
log f4 +129.95% 8.14% +109.30% +143.29% 50
log f8 +109.09% 4.19% +89.70% +114.54% 50
log f16 +119.16% 55.08% +5.73% +350.05% 50
log d2 -48.21% 9.50% -72.81% -13.93% 50
log d3 -65.73% 1.56% -72.87% -59.94% 50
log d4 -49.76% 2.24% -61.82% -45.95% 50
log d8 -49.66% 1.76% -61.43% -47.78% 50
log10 f2 +36.09% 2.80% +27.93% +42.93% 50
log10 f3 +97.05% 12.11% +62.51% +136.85% 50
log10 f4 +245.78% 15.49% +204.38% +278.37% 50
log10 f8 +187.55% 33.31% +15.03% +224.13% 50
log10 f16 +193.92% 29.42% +79.12% +259.62% 50
log10 d2 -48.23% 2.91% -55.03% -38.54% 50
log10 d3 -64.86% 11.78% -83.01% +13.09% 50
log10 d4 -50.59% 4.56% -79.43% -39.55% 50
log10 d8 -48.54% 2.46% -54.13% -36.72% 50
log1p f2 +120.27% 23.86% +54.72% +235.34% 50
log1p f3 +115.57% 24.45% +36.81% +164.47% 50
log1p f4 +211.69% 71.23% +50.01% +377.56% 50
log1p f8 +191.27% 14.99% +128.20% +247.37% 50
log1p f16 +197.19% 18.40% +74.54% +225.30% 50
log1p d2 -45.95% 3.45% -56.23% -30.65% 50
log1p d3 -61.14% 1.29% -67.99% -57.25% 50
log1p d4 -47.42% 8.68% -76.78% -31.19% 50
log1p d8 -46.16% 1.69% -55.96% -43.47% 50
log2 f2 +15.41% 3.49% +8.34% +30.07% 50
log2 f3 +52.12% 5.51% +38.41% +63.39% 50
log2 f4 +190.15% 27.79% +129.71% +293.45% 50
log2 f8 +163.93% 27.88% +62.80% +290.40% 50
log2 f16 +170.29% 49.44% +44.09% +285.61% 50
log2 d2 -47.10% 3.11% -60.56% -38.45% 50
log2 d3 -65.15% 1.17% -68.76% -59.02% 50
log2 d4 -48.17% 1.19% -50.05% -41.72% 50
log2 d8 -45.84% 2.20% -49.82% -36.07% 50
max i2i1 -30.15% 1.68% -36.15% -25.76% 50
max i2i2 -28.84% 15.49% -84.33% +10.59% 50
max i3i1 +48.45% 32.12% +21.77% +254.09% 50
max i3i3 +44.56% 17.91% -39.14% +66.82% 50
max i4i1 +6.22% 66.78% -63.56% +246.74% 50
max i4i4 +2.15% 24.70% -50.73% +103.70% 50
max u2u1 -32.89% 13.83% -70.68% +51.49% 50
max u2u2 -7.20% 68.77% -76.15% +372.39% 50
max u3u1 +31.38% 11.41% -15.94% +58.01% 50
max u3u3 +37.64% 10.79% +11.69% +82.97% 50
max u4u1 -20.15% 5.79% -32.73% -2.40% 50
max u4u4 -1.41% 8.92% -43.16% +28.47% 50
max f2f1 -26.43% 1.54% -29.00% -19.28% 50
max f2f2 -29.68% 15.77% -39.54% +74.74% 50
max f3f1 -23.39% 13.06% -57.62% +55.21% 50
max f3f3 -11.35% 15.77% -25.28% +95.94% 50
max f4f1 -20.97% 5.34% -35.74% -7.19% 50
max f4f4 -1.06% 6.62% -40.32% +5.59% 50
max d2d1 -19.89% 4.43% -33.81% -4.95% 50
max d2d2 -19.52% 3.38% -29.34% -10.58% 50
max d3d1 -43.66% 50.64% -97.52% +160.27% 50
max d3d3 -43.53% 6.70% -74.73% -22.33% 50
max d4d1 -25.60% 60.92% -68.34% +341.38% 50
max d4d4 -1.78% 74.04% -75.55% +280.38% 50
min i2i1 +89.80% 578.03% -88.34% +3730.13% 50
min i2i2 +26.40% 149.78% -83.39% +603.90% 50
min i3i1 +193.48% 437.26% -91.95% +1946.93% 50
min i3i3 +275.90% 467.44% -83.21% +1966.12% 50
min i4i1 +16.35% 151.00% -91.74% +594.36% 50
min i4i4 +70.18% 260.78% -92.12% +1707.75% 50
min u2u1 -4.96% 138.11% -94.04% +764.64% 50
min u2u2 -6.85% 94.70% -95.44% +442.34% 50
min u3u1 +70.05% 107.58% -78.25% +460.32% 50
min u3u3 +47.40% 94.30% -80.76% +403.51% 50
min u4u1 +11.66% 116.23% -92.59% +567.36% 50
min u4u4 +48.12% 134.64% -75.77% +487.78% 50
min f2f1 -4.11% 135.08% -97.24% +785.28% 50
min f2f2 +23.07% 183.00% -95.44% +842.06% 50
min f3f1 -8.95% 85.51% -81.56% +366.88% 50
min f3f3 +0.45% 73.46% -89.64% +223.65% 50
min f4f1 +3.62% 94.78% -87.56% +405.71% 50
min f4f4 +35.03% 116.48% -85.47% +422.02% 50
min d2d1 -0.09% 114.84% -80.62% +752.71% 50
min d2d2 -7.24% 84.33% -82.80% +367.19% 50
min d3d1 -17.74% 116.84% -95.47% +420.26% 50
min d3d3 -42.92% 29.42% -73.98% +52.99% 50
min d4d1 -9.52% 93.23% -90.26% +482.69% 50
min d4d4 -16.32% 59.52% -92.90% +187.67% 50
mix f2f2f1 +75.83% 154.20% -82.65% +778.22% 50
mix f2f2f2 +79.89% 132.21% -86.84% +633.78% 50
mix f3f3f1 +140.28% 523.53% -84.98% +2653.74% 50
mix f3f3f3 +50.34% 131.06% -73.31% +460.05% 50
mix f4f4f1 +46.99% 136.43% -70.08% +689.83% 50
mix f4f4f4 +59.15% 110.52% -83.70% +426.64% 50
mix d2d2d1 +79.45% 259.93% -79.92% +1767.49% 50
mix d2d2d2 +87.01% 156.20% -73.16% +750.34% 50
mix d3d3d1 +24.52% 80.09% -67.73% +326.36% 50
mix d3d3d3 +14.86% 44.70% -66.28% +230.44% 50
mix d4d4d1 +43.60% 90.34% -71.27% +356.35% 50
mix d4d4d4 +47.03% 158.01% -86.55% +782.19% 50
nextafter f2f2 -28.80% 53.46% -89.38% +155.31% 50
nextafter f3f3 -26.20% 44.66% -78.09% +133.79% 50
nextafter f4f4 +3.87% 50.55% -67.98% +129.93% 50
nextafter f8f8 -23.16% 30.51% -71.50% +77.11% 50
nextafter f16f16 -13.61% 32.33% -59.40% +131.90% 50
nextafter d2d2 +33.16% 151.11% -88.28% +833.17% 50
nextafter d3d3 -32.10% 51.53% -87.28% +189.74% 50
nextafter d4d4 -19.01% 39.84% -72.10% +120.56% 50
nextafter d8d8 -19.98% 27.31% -71.22% +47.33% 50
norm_inf f2 +108.05% 129.80% -84.01% +537.14% 50
norm_inf f3 +217.27% 183.12% -44.10% +1037.33% 50
norm_inf f4 +224.78% 145.23% -33.22% +770.46% 50
norm_inf d2 +187.44% 298.98% -64.77% +1548.25% 50
norm_inf d3 -6.11% 96.26% -88.32% +536.82% 50
norm_inf d4 +99.29% 143.65% -64.66% +644.82% 50
norm_one f2 -28.16% 206.33% -91.49% +1407.97% 50
norm_one f3 -27.43% 82.61% -94.95% +469.24% 50
norm_one f4 +34.87% 161.74% -86.85% +635.97% 50
norm_one d2 -24.53% 98.64% -89.16% +544.10% 50
norm_one d3 -69.41% 26.86% -93.53% +34.70% 50
norm_one d4 -16.99% 94.86% -83.84% +466.30% 50
normalize f2 +16.41% 102.97% -89.37% +439.56% 50
normalize f3 +38.76% 164.83% -75.14% +770.11% 50
normalize f4 +15.43% 88.63% -88.29% +342.51% 50
normalize d2 +8.51% 98.87% -83.60% +361.11% 50
normalize d3 -19.85% 51.11% -80.44% +141.64% 50
normalize d4 +19.18% 80.41% -80.61% +269.67% 50
pow f2f2 +28.53% 79.08% -77.20% +294.28% 50
pow f3f3 +132.25% 101.84% -19.16% +531.88% 50
pow f4f4 +329.50% 175.49% +24.74% +741.30% 50
pow f8f8 +328.66% 186.61% +33.18% +776.76% 50
pow f16f16 +212.67% 74.79% +34.83% +534.01% 50
pow d2d2 -5.80% 29.68% -50.48% +77.78% 50
pow d3d3 -39.99% 14.85% -64.18% +11.79% 50
pow d4d4 -16.02% 20.79% -61.06% +49.50% 50
pow d8d8 -18.24% 16.10% -52.45% +54.14% 50
project f2f2 +18.20% 73.80% -81.03% +282.76% 50
project f3f3 +49.75% 143.11% -58.78% +716.00% 50
project f4f4 +99.69% 340.44% -92.23% +1678.02% 50
project d2d2 +67.21% 236.96% -86.04% +1235.81% 50
project d3d3 -35.32% 44.20% -83.17% +102.15% 50
project d4d4 +5.38% 97.99% -88.36% +514.58% 50
recip f1 -39.55% 40.76% -94.48% +165.43% 50
recip f2 -17.77% 84.45% -89.51% +335.49% 50
recip f3 -4.99% 114.67% -85.94% +719.67% 50
recip f4 +15.34% 52.54% -76.68% +235.74% 50
recip d1 -24.76% 56.01% -90.22% +316.55% 50
recip d2 -1.31% 105.19% -78.45% +619.20% 50
recip d3 -38.50% 77.22% -87.96% +370.69% 50
recip d4 +42.01% 127.26% -87.88% +641.62% 50
reduce_add i2 -3.18% 95.58% -94.34% +391.25% 50
reduce_add i3 -19.46% 61.81% -91.00% +248.48% 50
reduce_add i4 +40.85% 218.53% -92.63% +1180.02% 50
reduce_add u2 +5.00% 86.98% -90.63% +346.89% 50
reduce_add u3 +29.24% 154.64% -85.30% +848.33% 50
reduce_add u4 +9.48% 75.78% -79.55% +293.59% 50
reduce_add f2 -7.02% 115.06% -66.48% +781.19% 50
reduce_add f3 +22.99% 216.73% -95.63% +1196.66% 50
reduce_add f4 +11.31% 158.02% -94.44% +1009.22% 50
reduce_add d2 +9.30% 140.01% -90.95% +678.40% 50
reduce_add d3 -8.69% 140.71% -93.72% +760.27% 50
reduce_add d4 -3.55% 162.13% -92.84% +843.16% 50
reduce_max i2 -1.35% 133.67% -91.70% +670.13% 50
reduce_max i3 +29.20% 104.37% -86.70% +344.12% 50
reduce_max i4 +3.43% 120.75% -88.16% +578.79% 50
reduce_max u2 -17.62% 79.24% -92.42% +375.51% 50
reduce_max u3 -1.90% 104.11% -91.12% +408.41% 50
reduce_max u4 +9.94% 177.97% -86.06% +925.93% 50
reduce_max f2 +237.44% 266.25% -55.96% +1191.32% 50
reduce_max f3 +264.94% 205.81% -61.42% +903.43% 50
reduce_max f4 +398.90% 278.09% -31.01% +1528.74% 50
reduce_max d2 +113.08% 161.01% -79.68% +650.23% 50
reduce_max d3 +129.14% 116.85% -71.33% +596.66% 50
reduce_max d4 +233.39% 153.10% +3.14% +751.63% 50
reduce_min i2 -33.05% 50.45% -88.38% +288.37% 50
reduce_min i3 -13.36% 59.19% -79.48% +318.17% 50
reduce_min i4 -23.15% 34.56% -75.73% +181.75% 50
reduce_min u2 -34.45% 43.84% -95.34% +174.66% 50
reduce_min u3 +117.84% 482.84% -93.08% +2718.12% 50
reduce_min u4 +54.10% 367.89% -91.54% +1988.27% 50
reduce_min f2 +256.02% 339.78% -90.52% +1544.68% 50
reduce_min f3 +392.05% 699.76% -71.59% +4859.39% 50
reduce_min f4 +352.07% 261.05% -25.84% +1305.99% 50
reduce_min d2 +176.12% 203.63% -78.01% +949.14% 50
reduce_min d3 +143.58% 151.41% -57.45% +643.31% 50
reduce_min d4 +235.38% 212.73% -43.89% +1090.68% 50
reflect f2f2 +26.57% 102.41% -79.46% +359.26% 50
reflect f3f3 +31.99% 113.68% -74.47% +515.36% 50
reflect f4f4 +15.04% 91.37% -84.62% +442.31% 50
reflect d2d2 +18.29% 77.35% -86.77% +291.98% 50
reflect d3d3 -27.43% 64.77% -80.94% +264.03% 50
reflect d4d4 -24.59% 35.84% -89.05% +107.36% 50
refract f2f2f1 +57.31% 146.92% -63.37% +618.56% 50
refract f3f3f1 +48.89% 148.84% -81.92% +708.32% 50
refract f4f4f1 +52.36% 153.06% -88.16% +658.65% 50
refract d2d2d1 +24.76% 101.45% -76.15% +475.97% 50
refract d3d3d1 -23.64% 43.76% -76.54% +150.08% 50
refract d4d4d1 +11.27% 89.02% -67.55% +404.11% 50
remainder f2f2 -71.26% 1.97% -76.11% -64.33% 50
remainder f3f3 -58.44% 4.29% -76.24% -48.36% 50
remainder f4f4 -24.00% 4.88% -47.75% -12.85% 50
remainder f8f8 -49.85% 3.94% -66.20% -40.63% 50
remainder f16f16 -46.53% 4.99% -73.25% -32.07% 50
remainder d2d2 -17.96% 3.94% -22.18% +1.31% 50
remainder d3d3 -60.17% 3.89% -83.79% -56.02% 50
remainder d4d4 -42.29% 5.49% -57.32% -21.03% 50
remainder d8d8 -45.18% 9.88% -68.41% -17.28% 50
round f2 -55.99% 2.78% -59.64% -40.23% 50
round f3 -33.44% 4.96% -48.43% -11.51% 50
round f4 -35.77% 3.05% -42.49% -22.03% 50
round f8 -39.00% 7.00% -67.25% -28.11% 50
round f16 -34.86% 3.07% -41.41% -22.90% 50
round d2 -39.23% 3.14% -48.66% -30.34% 50
round d3 -57.73% 1.95% -62.14% -52.22% 50
round d4 -48.85% 3.95% -65.87% -34.77% 50
round d8 -34.92% 5.38% -46.90% -5.90% 50
rsqrt f1 -1.66% 3.08% -13.12% +12.11% 50
rsqrt f2 -1.69% 1.51% -6.26% +1.50% 50
rsqrt f3 +111.24% 7.08% +73.75% +133.56% 50
rsqrt f4 +2.94% 2.77% -9.56% +6.84% 50
rsqrt d1 -47.10% 2.05% -54.76% -37.85% 50
rsqrt d2 +0.21% 3.25% -15.92% +10.23% 50
rsqrt d3 -39.32% 1.87% -41.67% -28.94% 50
rsqrt d4 -0.35% 4.13% -7.00% +18.71% 50
sign f1 -29.29% 5.23% -52.27% -16.25% 50
sign f2 +23.66% 7.38% -2.55% +44.77% 50
sign f3 +15.52% 9.15% -13.20% +42.87% 50
sign f4 +48.22% 14.31% +1.91% +92.03% 50
sign d1 -27.33% 9.90% -56.75% +15.09% 50
sign d2 +17.76% 21.79% -8.23% +159.79% 50
sign d3 -36.11% 21.73% -67.09% +47.66% 50
sign d4 +20.38% 46.74% -74.69% +182.98% 50
simd_abs i2 -5.54% 65.92% -64.61% +396.45% 50
simd_abs i3 -22.23% 8.49% -48.99% +6.90% 50
simd_abs i4 -9.98% 9.88% -40.58% +19.65% 50
simd_abs f2 -20.25% 3.01% -31.66% -11.38% 50
simd_abs f3 -20.53% 5.00% -39.09% -4.77% 50
simd_abs f4 -23.63% 2.71% -33.85% -17.31% 50
simd_abs f8 -13.97% 5.07% -25.79% +11.84% 50
simd_abs f16 -14.93% 2.75% -24.84% -8.96% 50
simd_abs d2 -20.63% 4.76% -29.99% +4.03% 50
simd_abs d3 -48.61% 3.35% -53.63% -35.10% 50
simd_abs d4 -26.09% 11.15% -45.05% +29.66% 50
simd_abs d8 -13.66% 4.46% -19.17% +2.17% 50
simd_clamp i2i2i2 -23.88% 8.28% -38.78% +22.02% 50
simd_clamp i3i3i3 -1.79% 11.01% -39.23% +39.11% 50
simd_clamp i4i4i4 +0.25% 6.12% -31.58% +10.00% 50
simd_clamp u2u2u2 -29.29% 5.00% -40.51% -7.07% 50
simd_clamp u3u3u3 -8.14% 8.03% -36.89% +9.13% 50
simd_clamp u4u4u4 +0.20% 7.74% -22.83% +17.05% 50
simd_clamp f1f1f1 -24.69% 2.77% -38.92% -21.26% 50
simd_clamp f2f2f2 -28.76% 3.93% -47.92% -12.57% 50
simd_clamp f3f3f3 -2.90% 7.23% -20.21% +29.70% 50
simd_clamp f4f4f4 +8.25% 65.49% -89.67% +380.75% 50
simd_clamp f8f8f8 -25.26% 9.16% -59.21% +7.18% 50
simd_clamp f16f16f16 -16.22% 20.31% -57.87% +55.72% 50
simd_clamp d1d1d1 -10.56% 79.06% -82.27% +451.47% 50
simd_clamp d2d2d2 -15.49% 14.57% -79.56% +30.61% 50
simd_clamp d3d3d3 -34.04% 12.39% -45.35% +45.66% 50
simd_clamp d4d4d4 -24.71% 4.65% -34.66% -4.07% 50
simd_clamp d8d8d8 -20.17% 3.41% -29.29% -7.57% 50
simd_cross f2f2 -20.49% 4.71% -32.19% -3.13% 50
simd_cross f3f3 -4.19% 11.02% -37.10% +33.91% 50
simd_cross d2d2 -40.47% 7.28% -55.64% -13.17% 50
simd_cross d3d3 -36.85% 5.06% -50.58% -23.43% 50
simd_distance_squared f2f2 +2.16% 5.58% -14.97% +23.57% 50
simd_distance_squared f3f3 +4.84% 6.94% -27.96% +13.26% 50
simd_distance_squared f4f4 +8.22% 11.33% -19.32% +54.82% 50
simd_distance_squared f8f8 -17.89% 8.47% -32.39% +33.68% 50
simd_distance_squared f16f16 -10.22% 8.42% -53.90% +7.64% 50
simd_distance_squared d2d2 -7.78% 12.51% -47.32% +30.34% 50
simd_distance_squared d3d3 -24.80% 4.88% -29.13% -6.73% 50
simd_distance_squared d4d4 -27.02% 4.24% -39.68% -15.77% 50
simd_distance_squared d8d8 -26.58% 12.48% -46.19% +17.60% 50
simd_dot f2f2 -0.18% 9.56% -25.87% +29.92% 50
simd_dot f3f3 +1.06% 6.10% -21.43% +23.05% 50
simd_dot f4f4 +14.68% 30.22% -23.39% +161.28% 50
simd_dot f8f8 -20.07% 11.22% -52.24% +14.62% 50
simd_dot f16f16 -4.47% 28.92% -59.08% +116.23% 50
simd_dot d2d2 -6.38% 32.98% -82.73% +88.59% 50
simd_dot d3d3 -5.94% 80.87% -76.09% +502.90% 50
simd_dot d4d4 -23.29% 5.71% -42.38% -6.51% 50
simd_dot d8d8 -30.04% 4.91% -36.29% -4.46% 50
simd_equal i2i2 +36.01% 94.92% -88.71% +669.89% 50
simd_equal i3i3 +83.05% 23.26% -5.29% +155.35% 50
simd_equal i4i4 -3.59% 13.52% -60.92% +43.19% 50
simd_equal u2u2 +26.52% 9.26% -1.06% +71.17% 50
simd_equal u3u3 +81.84% 15.14% +10.09% +115.72% 50
simd_equal u4u4 -9.73% 5.43% -24.68% +2.99% 50
simd_equal f2f2 +0.95% 18.89% -42.47% +84.62% 50
simd_equal f3f3 -17.31% 7.70% -27.36% +17.87% 50
simd_equal f4f4 -19.00% 3.43% -34.59% -6.42% 50
simd_equal f8f8 -31.88% 3.15% -39.24% -23.12% 50
simd_equal f16f16 -17.29% 3.56% -32.86% -4.67% 50
simd_equal d2d2 +21.50% 5.13% +1.98% +43.71% 50
simd_equal d3d3 -38.67% 3.13% -50.75% -26.75% 50
simd_equal d4d4 -34.20% 1.60% -37.68% -30.47% 50
simd_equal d8d8 -31.12% 3.06% -36.37% -17.13% 50
simd_fast_distance f2f2 +1.81% 4.67% -9.17% +20.58% 50
simd_fast_distance f3f3 +27.55% 6.09% -3.19% +54.74% 50
simd_fast_distance f4f4 +3.58% 4.74% -11.77% +23.19% 50
simd_fast_distance f8f8 -8.89% 3.26% -19.72% +0.22% 50
simd_fast_distance f16f16 -1.94% 7.23% -17.67% +31.17% 50
simd_fast_distance d2d2 -7.23% 11.04% -36.95% +33.61% 50
simd_fast_distance d3d3 -17.73% 33.31% -64.35% +93.06% 50
simd_fast_distance d4d4 -22.27% 23.35% -75.28% +39.36% 50
simd_fast_distance d8d8 -23.93% 24.14% -56.95% +122.12% 50
simd_fast_length f2 +8.01% 6.87% -16.62% +26.25% 50
simd_fast_length f3 +8.07% 8.25% -17.69% +40.23% 50
simd_fast_length f4 +40.10% 9.50% +10.52% +75.01% 50
simd_fast_length f8 +10.00% 5.65% -13.97% +27.32% 50
simd_fast_length f16 +19.88% 8.80% +1.32% +55.35% 50
simd_fast_length d2 -7.14% 3.14% -19.60% +1.18% 50
simd_fast_length d3 -26.47% 3.38% -42.14% -20.33% 50
simd_fast_length d4 -17.10% 3.99% -26.64% -3.46% 50
simd_fast_length d8 -18.26% 4.87% -39.78% -0.51% 50
simd_fast_normalize f2 +16.96% 6.76% -13.63% +36.59% 50
simd_fast_normalize f3 +29.46% 14.58% -50.82% +42.94% 50
simd_fast_normalize f4 +33.78% 15.04% +0.51% +121.06% 50
simd_fast_normalize f8 +14.07% 15.76% -27.68% +52.66% 50
simd_fast_normalize f16 +23.30% 9.21% +0.13% +47.30% 50
simd_fast_normalize d2 -20.62% 4.03% -42.34% -6.64% 50
simd_fast_normalize d3 -33.13% 5.70% -42.90% -14.36% 50
simd_fast_normalize d4 +2.35% 8.38% -22.60% +42.93% 50
simd_fast_normalize d8 -13.45% 3.81% -23.55% -1.33% 50
simd_fast_project f2f2 +12.95% 8.94% -24.84% +32.07% 50
simd_fast_project f3f3 +17.27% 9.87% -16.95% +41.68% 50
simd_fast_project f4f4 +36.80% 33.76% -44.37% +194.75% 50
simd_fast_project f8f8 +3.43% 30.88% -88.59% +126.70% 50
simd_fast_project f16f16 +18.14% 38.51% -55.45% +127.28% 50
simd_fast_project d2d2 +19.24% 151.25% -68.95% +1070.05% 50
simd_fast_project d3d3 -31.26% 8.81% -47.21% +9.14% 50
simd_fast_project d4d4 -16.87% 6.16% -28.16% +9.49% 50
simd_fast_project d8d8 -17.31% 2.90% -21.80% -4.06% 50
simd_fast_recip f1 -0.80% 6.33% -20.90% +33.00% 50
simd_fast_recip f2 -18.19% 5.04% -37.79% +7.90% 50
simd_fast_recip f3 -12.18% 4.69% -25.96% -2.80% 50
simd_fast_recip f4 -0.52% 6.87% -24.17% +21.10% 50
simd_fast_recip f8 -14.27% 7.78% -28.88% +23.79% 50
simd_fast_recip f16 -18.03% 7.58% -36.47% +7.36% 50
simd_fast_recip d1 -24.28% 10.59% -58.21% -3.07% 50
simd_fast_recip d2 -17.07% 20.17% -45.96% +89.67% 50
simd_fast_recip d3 -49.28% 2.36% -56.75% -43.39% 50
simd_fast_recip d4 -9.25% 8.37% -28.02% +24.77% 50
simd_fast_recip d8 -18.68% 11.20% -81.52% +0.41% 50
simd_fast_rsqrt f1 +27.41% 16.61% -59.03% +61.92% 50
simd_fast_rsqrt f2 +13.60% 8.14% -3.20% +47.70% 50
simd_fast_rsqrt f3 +121.63% 13.61% +47.21% +135.31% 50
simd_fast_rsqrt f4 +4.80% 9.66% -18.68% +44.41% 50
simd_fast_rsqrt f8 -8.48% 8.68% -33.73% +25.41% 50
simd_fast_rsqrt f16 -4.06% 41.82% -39.53% +242.74% 50
simd_fast_rsqrt d1 -28.62% 57.62% -74.24% +255.23% 50
simd_fast_rsqrt d2 +4.60% 40.41% -76.87% +145.31% 50
simd_fast_rsqrt d3 -30.60% 42.58% -74.22% +231.33% 50
simd_fast_rsqrt d4 -0.76% 8.35% -8.59% +36.84% 50
simd_fast_rsqrt d8 -1.04% 4.90% -9.95% +17.99% 50
simd_fract f1 +32.41% 22.62% -39.16% +101.44% 50
simd_fract f2 +20.73% 7.23% +0.99% +38.58% 50
simd_fract f3 +9.06% 20.53% -33.51% +108.86% 50
simd_fract f4 +33.88% 11.82% -11.99% +84.70% 50
simd_fract f8 +18.55% 10.38% -1.40% +73.04% 50
simd_fract f16 +34.66% 13.91% -35.47% +82.52% 50
simd_fract d1 -32.70% 6.20% -43.59% -6.20% 50
simd_fract d2 -0.70% 5.32% -18.50% +14.48% 50
simd_fract d3 -38.17% 6.37% -51.73% -15.94% 50
simd_fract d4 +0.07% 16.15% -19.11% +98.56% 50
simd_fract d8 -9.03% 6.71% -21.94% +16.34% 50
simd_incircle f2f2f2f2 -39.17% 4.36% -46.86% -26.26% 50
simd_incircle d2d2d2d2 -36.17% 4.71% -42.55% -6.34% 50
simd_insphere f3f3f3f3f3 +67.03% 40.90% -17.91% +207.43% 50
simd_insphere d3d3d3d3d3 +1.18% 4.85% -19.37% +17.99% 50
simd_length_squared f2 -12.98% 4.49% -33.18% +5.83% 50
simd_length_squared f3 -6.67% 7.46% -24.91% +23.05% 50
simd_length_squared f4 +8.74% 9.38% -39.31% +21.14% 50
simd_length_squared f8 -15.07% 3.51% -27.68% -3.79% 50
simd_length_squared f16 +2.03% 6.07% -3.79% +35.57% 50
simd_length_squared d2 -8.82% 6.07% -19.56% +26.53% 50
simd_length_squared d3 -30.52% 2.37% -38.31% -24.00% 50
simd_length_squared d4 -20.50% 3.19% -34.34% -10.16% 50
simd_length_squared d8 -27.19% 7.75% -37.92% +21.80% 50
simd_max i2i2 -10.56% 25.66% -39.72% +138.15% 50
simd_max i3i3 +48.82% 11.26% +22.07% +77.92% 50
simd_max i4i4 -6.43% 4.75% -22.11% +8.07% 50
simd_max u2u2 -14.34% 3.58% -24.12% -7.06% 50
simd_max u3u3 +29.80% 8.36% +11.28% +74.24% 50
simd_max u4u4 -5.53% 3.99% -10.70% +12.25% 50
simd_max f1f1 -30.49% 2.17% -36.22% -21.28% 50
simd_max f2f2 -13.47% 12.50% -39.20% +33.17% 50
simd_max f3f3 -10.11% 24.12% -34.34% +148.89% 50
simd_max f4f4 -3.01% 17.49% -77.71% +57.15% 50
simd_max f8f8 -13.38% 42.80% -79.70% +212.82% 50
simd_max f16f16 -11.97% 47.68% -67.45% +255.35% 50
simd_max d1d1 -23.48% 13.50% -42.03% +65.11% 50
simd_max d2d2 -16.61% 9.35% -34.94% +7.53% 50
simd_max d3d3 -41.54% 4.74% -47.77% -23.89% 50
simd_max d4d4 -21.18% 6.66% -31.88% -1.28% 50
simd_max d8d8 -17.86% 5.28% -26.62% +4.97% 50
simd_min i2i2 -14.07% 8.16% -33.83% +17.38% 50
simd_min i3i3 +52.20% 33.04% -22.39% +155.08% 50
simd_min i4i4 -6.02% 13.68% -43.38% +51.45% 50
simd_min u2u2 -11.17% 14.86% -27.39% +77.72% 50
simd_min u3u3 +29.01% 12.17% -4.89% +66.80% 50
simd_min u4u4 -4.07% 9.57% -27.34% +32.13% 50
simd_min f1f1 -30.61% 10.48% -55.33% +27.09% 50
simd_min f2f2 -14.79% 3.30% -25.73% -5.38% 50
simd_min f3f3 -14.25% 8.56% -68.50% +0.75% 50
simd_min f4f4 -6.03% 3.77% -18.27% +11.38% 50
simd_min f8f8 -25.04% 3.13% -31.28% -17.00% 50
simd_min f16f16 -18.85% 1.93% -22.96% -13.68% 50
simd_min d1d1 -25.78% 7.00% -47.15% +8.60% 50
simd_min d2d2 -18.01% 9.65% -55.64% +6.37% 50
simd_min d3d3 -39.30% 23.88% -73.45% +83.49% 50
simd_min d4d4 -16.17% 29.15% -81.71% +69.56% 50
simd_min d8d8 -16.00% 26.18% -53.35% +130.64% 50
simd_mix f1f1f1 -23.54% 3.36% -37.87% -15.86% 50
simd_mix f2f2f2 -1.36% 7.44% -20.95% +15.13% 50
simd_mix f3f3f3 +4.53% 7.60% -19.53% +24.02% 50
simd_mix f4f4f4 +6.39% 9.43% -12.02% +62.05% 50
simd_mix f8f8f8 -26.37% 4.45% -33.56% -8.12% 50
simd_mix f16f16f16 -16.79% 4.80% -29.16% +7.65% 50
simd_mix d1d1d1 -19.05% 18.35% -50.48% +92.14% 50
simd_mix d2d2d2 +9.41% 6.97% -7.36% +35.20% 50
simd_mix d3d3d3 -34.93% 3.47% -48.62% -20.70% 50
simd_mix d4d4d4 -21.10% 2.49% -28.87% -13.20% 50
simd_mix d8d8d8 -16.94% 2.73% -27.72% -9.64% 50
simd_norm_inf f2 +148.87% 11.89% +76.24% +166.01% 50
simd_norm_inf f3 +211.15% 6.84% +196.03% +228.30% 50
simd_norm_inf f4 +261.21% 12.58% +200.61% +305.17% 50
simd_norm_inf f8 +242.40% 17.40% +204.64% +322.84% 50
simd_norm_inf f16 +423.63% 182.16% +104.77% +1130.40% 50
simd_norm_inf d2 +133.19% 32.34% +20.04% +261.25% 50
simd_norm_inf d3 +94.85% 23.43% +12.34% +155.41% 50
simd_norm_inf d4 +129.75% 12.04% +99.39% +159.16% 50
simd_norm_inf d8 +126.95% 14.26% +90.47% +198.11% 50
simd_norm_one f2 -3.90% 10.17% -36.37% +30.98% 50
simd_norm_one f3 -5.06% 7.52% -20.75% +26.33% 50
simd_norm_one f4 +13.22% 5.92% -2.34% +33.37% 50
simd_norm_one f8 -16.64% 3.18% -24.78% +0.13% 50
simd_norm_one f16 +2.66% 5.25% -7.06% +28.95% 50
simd_norm_one d2 -10.33% 8.11% -51.84% +13.50% 50
simd_norm_one d3 -29.58% 4.77% -42.23% -7.41% 50
simd_norm_one d4 -18.85% 7.56% -26.39% +27.47% 50
simd_norm_one d8 -28.77% 6.22% -51.70% +1.28% 50
simd_orient f2f2 -81.14% 0.86% -83.47% -77.57% 50
simd_orient d2d2 -78.17% 3.92% -81.03% -51.34% 50
simd_orient f2f2f2 -76.36% 1.08% -78.08% -70.50% 50
simd_orient f3f3f3 -61.49% 1.60% -64.59% -54.97% 50
simd_orient d2d2d2 -70.32% 1.77% -72.10% -62.19% 50
simd_orient d3d3d3 -71.05% 1.79% -73.76% -64.40% 50
simd_orient f3f3f3f3 -51.59% 16.29% -78.29% +9.25% 50
simd_orient d3d3d3d3 -68.41% 1.76% -72.98% -60.03% 50
simd_precise_distance f2f2 -1.40% 17.64% -54.56% +89.42% 50
simd_precise_distance f3f3 +51.39% 79.82% -45.97% +414.89% 50
simd_precise_distance f4f4 +2.38% 15.58% -80.22% +44.88% 50
simd_precise_distance f8f8 -8.81% 7.04% -23.34% +24.54% 50
simd_precise_distance f16f16 -1.18% 2.38% -7.75% +8.33% 50
simd_precise_distance d2d2 -12.60% 5.05% -38.04% +3.47% 50
simd_precise_distance d3d3 -22.92% 4.30% -40.86% -9.64% 50
simd_precise_distance d4d4 -21.74% 2.84% -35.07% -11.85% 50
simd_precise_distance d8d8 -27.21% 3.12% -33.59% -17.58% 50
simd_precise_length f2 +11.94% 18.49% +1.50% +137.91% 50
simd_precise_length f3 +1.10% 3.22% -14.15% +9.30% 50
simd_precise_length f4 +44.29% 4.64% +14.73% +46.97% 50
simd_precise_length f8 +10.86% 3.11% -0.17% +19.76% 50
simd_precise_length f16 +14.10% 4.01% +0.30% +26.49% 50
simd_precise_length d2 -12.23% 9.29% -39.40% +39.99% 50
simd_precise_length d3 -26.45% 2.69% -35.72% -19.50% 50
simd_precise_length d4 -16.83% 7.33% -38.00% +7.67% 50
simd_precise_length d8 -18.62% 9.47% -49.36% +3.40% 50
simd_precise_normalize f2 -5.24% 18.56% -40.56% +87.17% 50
simd_precise_normalize f3 +15.54% 57.09% -70.83% +223.30% 50
simd_precise_normalize f4 +17.44% 96.84% -69.45% +637.90% 50
simd_precise_normalize f8 +17.03% 57.78% -69.70% +251.44% 50
simd_precise_normalize f16 +17.30% 10.70% -53.37% +29.30% 50
simd_precise_normalize d2 -22.20% 4.76% -49.34% -16.42% 50
simd_precise_normalize d3 -30.83% 3.04% -35.27% -23.56% 50
simd_precise_normalize d4 -0.66% 7.29% -32.47% +13.74% 50
simd_precise_normalize d8 -8.79% 4.24% -29.89% +2.63% 50
simd_precise_project f2f2 +19.78% 4.38% -4.60% +28.74% 50
simd_precise_project f3f3 +19.36% 3.28% +1.72% +28.04% 50
simd_precise_project f4f4 +18.35% 7.18% -4.82% +39.78% 50
simd_precise_project f8f8 +4.62% 9.48% -4.87% +48.56% 50
simd_precise_project f16f16 +18.68% 2.02% +13.94% +25.41% 50
simd_precise_project d2d2 -2.92% 3.68% -19.17% +5.10% 50
simd_precise_project d3d3 -31.84% 3.86% -39.92% -14.15% 50
simd_precise_project d4d4 -13.10% 5.83% -20.86% +19.39% 50
simd_precise_project d8d8 -10.41% 18.97% -57.81% +81.24% 50
simd_precise_recip f1 -3.76% 54.01% -56.44% +348.14% 50
simd_precise_recip f2 -22.38% 19.36% -63.25% +66.75% 50
simd_precise_recip f3 -23.60% 21.48% -70.87% +60.09% 50
simd_precise_recip f4 +10.38% 42.92% -79.08% +281.81% 50
simd_precise_recip f8 -10.57% 3.72% -17.80% +2.08% 50
simd_precise_recip f16 -11.26% 6.34% -26.81% +21.04% 50
simd_precise_recip d1 -24.36% 4.19% -46.25% -7.03% 50
simd_precise_recip d2 -19.09% 3.52% -32.80% -6.37% 50
simd_precise_recip d3 -47.53% 3.77% -52.27% -30.52% 50
simd_precise_recip d4 -6.53% 2.69% -12.48% +2.88% 50
simd_precise_recip d8 -13.46% 1.92% -20.11% -8.50% 50
simd_precise_rsqrt f1 +2.01% 5.10% -3.29% +35.27% 50
simd_precise_rsqrt f2 -2.88% 4.70% -10.33% +23.19% 50
simd_precise_rsqrt f3 +93.17% 7.49% +65.26% +110.82% 50
simd_precise_rsqrt f4 +3.16% 12.36% -51.81% +45.60% 50
simd_precise_rsqrt f8 -8.91% 4.75% -23.13% +13.60% 50
simd_precise_rsqrt f16 -11.31% 1.90% -14.70% -4.26% 50
simd_precise_rsqrt d1 -42.91% 1.80% -51.24% -38.23% 50
simd_precise_rsqrt d2 +1.93% 5.58% -7.92% +25.15% 50
simd_precise_rsqrt d3 -37.18% 2.71% -44.65% -28.99% 50
simd_precise_rsqrt d4 -0.66% 8.38% -39.10% +19.81% 50
simd_precise_rsqrt d8 -1.61% 13.52% -22.86% +66.66% 50
simd_reflect f2f2 +13.44% 68.96% -68.89% +309.13% 50
simd_reflect f3f3 +33.40% 69.75% -58.65% +423.17% 50
simd_reflect f4f4 +10.10% 60.39% -69.74% +318.52% 50
simd_reflect d2d2 -0.32% 4.90% -13.92% +15.26% 50
simd_reflect d3d3 -36.76% 4.15% -44.81% -18.74% 50
simd_reflect d4d4 -18.67% 4.06% -30.19% -2.66% 50
simd_refract f2f2f1 +4.88% 4.26% -14.41% +21.89% 50
simd_refract f3f3f1 +6.80% 3.85% -8.69% +20.24% 50
simd_refract f4f4f1 +3.59% 3.21% -4.38% +13.43% 50
simd_refract d2d2d1 -4.10% 2.93% -16.21% +0.90% 50
simd_refract d3d3d1 -25.79% 7.62% -40.74% +15.47% 50
simd_refract d4d4d1 -16.09% 6.19% -50.54% -1.79% 50
simd_sign f1 -17.27% 5.02% -34.18% -1.37% 50
simd_sign f2 +21.23% 5.78% +0.57% +43.14% 50
simd_sign f3 +20.02% 18.51% -27.30% +125.78% 50
simd_sign f4 +42.07% 10.56% +21.32% +89.93% 50
simd_sign f8 +52.14% 23.16% +16.09% +132.91% 50
simd_sign f16 +89.13% 5.76% +73.61% +104.84% 50
simd_sign d1 -20.76% 5.30% -52.23% -7.44% 50
simd_sign d2 +5.05% 7.17% -6.82% +43.83% 50
simd_sign d3 -37.85% 3.22% -44.48% -26.71% 50
simd_sign d4 +11.76% 8.52% -5.63% +42.23% 50
simd_sign d8 +14.37% 33.27% -53.34% +172.23% 50
simd_smoothstep f1f1f1 +254.50% 122.77% +0.32% +833.46% 50
simd_smoothstep f2f2f2 +340.69% 20.61% +238.11% +395.26% 50
simd_smoothstep f3f3f3 +28.18% 6.34% +2.32% +52.89% 50
simd_smoothstep f4f4f4 +24.61% 9.64% -3.91% +76.97% 50
simd_smoothstep f8f8f8 +36.93% 3.53% +28.05% +51.01% 50
simd_smoothstep f16f16f16 +38.67% 4.34% +23.90% +54.33% 50
simd_smoothstep d1d1d1 +225.67% 17.92% +157.71% +278.74% 50
simd_smoothstep d2d2d2 +6.48% 7.00% -25.33% +32.84% 50
simd_smoothstep d3d3d3 -31.72% 2.64% -34.91% -22.10% 50
simd_smoothstep d4d4d4 -18.01% 2.92% -24.20% -10.49% 50
simd_smoothstep d8d8d8 -9.23% 2.25% -14.44% -1.55% 50
simd_step f1f1 -28.20% 8.73% -42.52% +21.34% 50
simd_step f2f2 -22.03% 20.45% -70.15% +82.65% 50
simd_step f3f3 +4.91% 14.57% -26.91% +78.49% 50
simd_step f4f4 -0.59% 6.66% -22.43% +28.68% 50
simd_step f8f8 -13.85% 25.87% -65.22% +88.31% 50
simd_step f16f16 -16.09% 21.95% -62.48% +63.35% 50
simd_step d1d1 -28.12% 6.40% -57.29% -7.56% 50
simd_step d2d2 -20.64% 6.06% -49.34% -3.41% 50
simd_step d3d3 -36.52% 4.48% -50.66% -20.41% 50
simd_step d4d4 -21.69% 2.99% -29.04% -13.09% 50
simd_step d8d8 -19.22% 3.56% -28.56% -9.01% 50
sin f2 +238.95% 8.52% +210.60% +270.13% 50
sin f3 +393.46% 13.51% +353.58% +432.43% 50
sin f4 +634.85% 31.26% +487.36% +666.56% 50
sin f8 +548.66% 11.84% +496.17% +571.27% 50
sin f16 +586.37% 94.00% +251.55% +883.00% 50
sin d2 +257.43% 9.53% +203.34% +273.76% 50
sin d3 +146.07% 4.33% +127.76% +158.43% 50
sin d4 +235.69% 6.12% +220.53% +256.15% 50
sin d8 +255.99% 9.85% +215.62% +294.48% 50
sinh f2 +51.96% 42.32% -42.61% +169.30% 50
sinh f3 +147.38% 110.84% -46.41% +629.37% 50
sinh f4 +206.01% 55.67% +80.08% +539.64% 50
sinh f8 +161.53% 17.05% +106.94% +266.24% 50
sinh f16 +162.96% 9.10% +125.72% +182.22% 50
sinh d2 +167.42% 11.84% +121.59% +208.01% 50
sinh d3 +62.93% 4.70% +45.76% +75.79% 50
sinh d4 +126.67% 3.43% +116.32% +134.13% 50
sinh d8 +144.09% 22.95% +78.59% +228.49% 50
sinpi f2 +249.41% 79.66% +60.72% +462.45% 50
sinpi f3 +408.49% 17.00% +345.10% +454.40% 50
sinpi f4 +642.54% 36.62% +444.87% +719.66% 50
sinpi f8 +542.20% 43.21% +471.12% +737.96% 50
sinpi f16 +515.19% 14.43% +464.96% +583.08% 50
sinpi d2 +182.20% 25.07% +115.87% +328.76% 50
sinpi d3 +75.57% 45.18% -29.83% +322.37% 50
sinpi d4 +164.79% 83.69% -35.52% +425.66% 50
sinpi d8 +150.39% 16.00% +78.67% +183.14% 50
smoothstep f2f2f2 +354.21% 27.21% +219.53% +427.86% 50
smoothstep f3f3f3 +32.98% 9.47% -6.06% +76.98% 50
smoothstep f4f4f4 +17.10% 21.72% -84.85% +40.34% 50
smoothstep d2d2d2 +5.86% 4.65% -9.70% +14.86% 50
smoothstep d3d3d3 -29.55% 8.31% -33.69% +25.00% 50
smoothstep d4d4d4 -18.58% 2.76% -25.25% -8.85% 50
step f1f1 -30.28% 1.91% -32.90% -20.98% 50
step f2f2 -27.19% 3.04% -33.09% -11.06% 50
step f3f3 +17.20% 76.99% -92.11% +499.71% 50
step f4f4 +10.15% 62.04% -70.06% +291.88% 50
step d1d1 -12.79% 81.12% -90.21% +493.49% 50
step d2d2 -20.31% 5.53% -36.94% +1.92% 50
step d3d3 -34.79% 4.89% -40.43% -11.05% 50
step d4d4 -25.10% 7.80% -40.14% +7.10% 50
tan f2 +213.06% 89.98% +72.70% +689.26% 50
tan f3 +370.34% 139.67% +76.60% +1045.83% 50
tan f4 +486.26% 12.97% +448.97% +524.39% 50
tan f8 +448.27% 9.62% +399.19% +469.71% 50
tan f16 +461.27% 14.17% +400.65% +479.08% 50
tan d2 +210.74% 29.86% +82.86% +286.21% 50
tan d3 +120.42% 77.78% -24.14% +517.16% 50
tan d4 +180.71% 75.98% -23.39% +567.26% 50
tan d8 +178.98% 11.96% +129.80% +220.10% 50
tanh f2 +35.19% 7.16% +18.17% +61.55% 50
tanh f3 +120.82% 10.82% +86.64% +169.96% 50
tanh f4 +195.26% 7.87% +166.55% +223.41% 50
tanh f8 +166.45% 5.40% +153.76% +196.17% 50
tanh f16 +192.70% 39.11% +53.10% +290.85% 50
tanh d2 +25.39% 3.44% +6.34% +33.67% 50
tanh d3 -15.61% 2.63% -23.56% -8.64% 50
tanh d4 +22.20% 2.57% +10.78% +26.65% 50
tanh d8 +25.13% 10.40% -12.39% +56.03% 50
tanpi f2 +218.76% 74.43% -5.53% +478.56% 50
tanpi f3 +351.65% 12.75% +320.05% +388.11% 50
tanpi f4 +603.05% 20.58% +500.18% +642.12% 50
tanpi f8 +521.80% 16.37% +448.82% +557.34% 50
tanpi f16 +516.45% 23.26% +415.79% +561.82% 50
tanpi d2 +190.83% 7.96% +160.16% +211.16% 50
tanpi d3 +86.38% 30.65% +3.48% +185.96% 50
tanpi d4 +172.89% 54.39% +0.74% +348.12% 50
tanpi d8 +161.81% 6.45% +127.83% +175.43% 50
tgamma f2 -16.38% 0.77% -19.33% -14.69% 50
tgamma f3 -6.84% 7.42% -28.61% +16.65% 50
tgamma f4 -1.18% 1.08% -4.48% +0.92% 50
tgamma f8 +1.12% 5.29% -14.79% +16.32% 50
tgamma f16 +1.89% 6.43% -26.74% +22.31% 50
tgamma d2 -0.17% 7.96% -20.16% +31.64% 50
tgamma d3 -9.08% 4.43% -33.92% +4.28% 50
tgamma d4 -4.14% 6.57% -32.24% +24.03% 50
tgamma d8 +0.74% 5.63% -9.98% +24.63% 50
trunc f2 -36.75% 3.18% -42.82% -18.80% 50
trunc f3 -24.81% 6.37% -49.80% +4.06% 50
trunc f4 -24.18% 4.19% -35.36% +0.54% 50
trunc d2 -24.24% 3.68% -35.05% -6.61% 50
trunc d3 -53.41% 5.60% -61.09% -19.40% 50
trunc d4 -31.50% 4.48% -45.66% -8.64% 50