I was surprised to see results that are significantly different (by much more than I would except when using normal ffast-math optimizations) when running julia with --math-mode=fast. It seems like a very rough approximation is being used for exp() ( 0.1% error for exp(1.0)). Is this documented an…

I don’t know, but try doing the @fastmath call first? Maybe the compiler knows the result should be the same and just reuses it?

This is old (before Julia 1.0), but I don’t know if the situation has changed. If not, that would explain the difference you observed: [image] How to tell VS Code to use the fast math option? VS Code You really don’t want to use --math-mode=fast. You’re telling the compil…

The beginning of the code for exp looks something like: const a = 369.3299304675746 const b = 6.755399441055744e15 function f(x::Float64) z = muladd(x, a, b) z -= b end Without --math-mode=fast this gives the result 369.0, with it it gives 369.3299304675746. Looking at the generated code,…

That seems unfortunate. Is there any way to add something like @nofastmath to julia’s implementation of exp()? I see a similar constant and muladd in the glibc implementation, but that generally won’t be compiled with -ffast-math. Also for some reason, julia’s implementation is 24% slower than gl…

Why are we allowing changing a muladd to a mul? That just seems totally broken.

[image] robsmith11: That seems unfortunate I mean, the whole point of fast-math is trading off speed with correctness. If fast-math was to give always the correct results, it wouldn’t be fast-math, it would be the standard way of doing math.

[image] giordano: I mean, the whole point of fast-math is trading off speed with correctness. If fast-math was to give always the correct results, it wouldn’t be fast-math, it would be the standard way of doing math. Right, but generally when I use fast-math, I can reason about my own code an…

[image] Oscar_Smith: Why are we allowing changing a muladd to a mul? That just seems totally broken. It simplifies z = x * a + b # z = muladd(x, a, b) z -= b to z = x * a

Oh, that actually should be fixable. Rewriting that to unsafe_trunc should retain almost all the performance and remove the ability of fastmath to mess it up.

More generally, however, I don’t think @fastmath should be applied to inlined code from other functions. There are just way too many pitfalls there. Basically, @fastmath is used by people who think “this code is insensitive to rearrangements that are exact in exact arithmentic (e.g. re-associatio…

What's going on with exp() and --math-mode=fast?

General Usage

heliosdrm July 14, 2021, 9:23am 3

This is old (before Julia 1.0), but I don’t know if the situation has changed. If not, that would explain the difference you observed:

Topic		Replies	Views
Fastmath error? New to Julia fast-math	2	506	July 16, 2021
@fastmath macro accuracy General Usage numerics , fast-math	26	11494	May 12, 2020
How to tell VS Code to use the fast math option? VS Code fast-math	30	4112	December 7, 2018
Slow exp()? Numerics fast-math	12	2182	August 17, 2018
@fastmath switched off in Julia 0.7-Dev, deliberately? Performance fast-math	7	1798	June 24, 2018

What's going on with exp() and --math-mode=fast?

Related topics