# Performance of Float32 exponential

**URL:** <https://discourse.julialang.org/t/performance-of-float32-exponential/32549>\
**Category:** Performance\
**Created:** [December 21, 2019, 11:40am UTC](https://discourse.julialang.org/t/performance-of-float32-exponential/32549 "2019-12-21T11:40:06Z")\
**Posts on this page:** 1\
**Showing post:** 3

<div class="post-metadata">

**Author:** ![baggepinnen](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/baggepinnen/32/693_2.png) [@baggepinnen](https://discourse.julialang.org/u/baggepinnen)\
**Post date:** [December 21, 2019, 12:23pm UTC](https://discourse.julialang.org/t/performance-of-float32-exponential/32549/3 "2019-12-21T12:23:25Z")

</div>

See some discussion in this answer and thread

> [@Fast logsumexp](https://discourse.julialang.org/t/fast-logsumexp/22827/4):
>
> LLVM. GCC will vectorize log/exp/sin/etc with the appropriate optimization flags, but LLVM needs those in addition to -fveclib=SVML or some other vector library. using SIMDPirates, SLEEFPirates, LoopVectorization function logsumexp\_simdpirates!(w::Vector{T},we) where T offset = maximum(w) N = length(w) sl = SIMDPirates.Vec{4,T}((0.,0.,0.,0.)) @inbounds @simd for i = 1:4:N wl = SIMDPirates.vload(SIMDPirates.Vec{4,T}, w, i) @pirate wel = SLEEFPirates.exp(…

---

_[View the full topic](https://discourse.julialang.org/t/performance-of-float32-exponential/32549)._