# Floating point optimizations
**URL:** https://discourse.julialang.org/t/floating-point-optimizations/89585
**Category:** General Usage
**Tags:** question, float
**Created:** [November 1, 2022, 1:21am UTC](https://discourse.julialang.org/t/floating-point-optimizations/89585 "2022-11-01T01:21:43Z")
**Posts on this page:** 6
**Page:** 1
### Author:  [@guilhermebodin](https://discourse.julialang.org/u/guilhermebodin)
#### Post date: [November 1, 2022, 1:21am UTC](https://discourse.julialang.org/t/floating-point-optimizations/89585/1 "2022-11-01T01:21:43Z")
Running julia with `--math-mode=ieee` is somehow equivalent to compiling a c++ project in MSVC with /fp:strict?
---
### Author:  [@jling](https://discourse.julialang.org/u/jling)
#### Post date: [November 1, 2022, 2:19am UTC](https://discourse.julialang.org/t/floating-point-optimizations/89585/2 "2022-11-01T02:19:42Z")
yes, it overrides all `@fastmath`
---
### Author:  [@Oscar\_Smith](https://discourse.julialang.org/u/Oscar_Smith)
#### Post date: [November 1, 2022, 2:28am UTC](https://discourse.julialang.org/t/floating-point-optimizations/89585/3 "2022-11-01T02:28:48Z")
Note that as of 1.8 (or 1.9 I forget which) `--math-mode=fast` has been disabled (it’s now a noop) and before then, you shouldn’t use it.
---
### Author:  [@guilhermebodin](https://discourse.julialang.org/u/guilhermebodin)
#### Post date: [November 1, 2022, 2:34am UTC](https://discourse.julialang.org/t/floating-point-optimizations/89585/4 "2022-11-01T02:34:06Z")
Thank you guys for the answers!
To be more specific, if I run the same computations on two different architectures, does --math-mode=ieee guarantee that I will have the same results?
---
### Author:  [@Oscar\_Smith](https://discourse.julialang.org/u/Oscar_Smith)
#### Post date: [November 1, 2022, 2:41am UTC](https://discourse.julialang.org/t/floating-point-optimizations/89585/5 "2022-11-01T02:41:13Z")
it does not. Some operations like `muladd` can be compiled in different ways (either a fma or a multiply and add), and things like multithreading or random number generation can also lead to different results between different computers or different runs on the same computer.
---
### Author:  [@mikmoore](https://discourse.julialang.org/u/mikmoore)
#### Post date: [November 1, 2022, 2:59pm UTC](https://discourse.julialang.org/t/floating-point-optimizations/89585/6 "2022-11-01T14:59:14Z")
IEEE754 compliance ensures that primitives such as `+`, `-`, `*`, `/`, `sqrt`, `rem`, `fma` return the exact same results and aren’t transformed or reassociated without an explicit annotation like `@fastmath` or `@simd` or the use of `muladd` (for example, `a*b` becoming `a*(1/b)`, `(a+b)-c` becoming `a+(b-c)`, or `a*b+c` becoming `fma(a,b,c)`). As others have said, `--math-mode=ieee` at the command line disables these annotations globally (except maybe `muladd`? not sure on that one).
IEEE mode does not guarantee that functions like `sum` won’t have input- or architecture-dependent association orders (although `foldl(+,itr)` should be a consistent but slower alternative) or that non-primitive functions like `exp`, `sin`, or `^` will return the same results.
So for low-level calculation with a very limited set of primitive operations, IEEE mode should provide reproducible results. For more elaborate calculations involving more complicated functions or array-level operations, it will likely not. For general scientific computing, you probably don’t care. If you’re trying to write a precision implementation of `(sin(x)-x)/x^3`, you probably do care.
The default behavior is to use IEEE754 compliant math. You shouldn’t need to use `--math-mode=ieee` except if you’ve used annotations but want to disable them because you’re worried they’re causing trouble. `@fastmath` can introduce all sorts of insidious corner-case bugs so I’ll generally recommend against it. Most of the boost that `@fastmath` can deliver can usually be accomplished by manually reorganizing your expression (although sometimes this can be tedious). `@simd` is safer because it has a narrower use case - mostly to permit a reduction (like you’d see inside a `sum` loop) to be reordered.