# Float32 performance fluctuating over time
**URL:** https://discourse.julialang.org/t/float32-performance-fluctuating-over-time/70206
**Category:** Performance
**Created:** [October 22, 2021, 11:54am UTC](https://discourse.julialang.org/t/float32-performance-fluctuating-over-time/70206 "2021-10-22T11:54:32Z")
**Posts on this page:** 2
**Page:** 1
### Author:  [@Chiil](https://discourse.julialang.org/u/Chiil)
#### Post date: [October 22, 2021, 11:54am UTC](https://discourse.julialang.org/t/float32-performance-fluctuating-over-time/70206/1 "2021-10-22T11:54:32Z")
I have coded a CFD code in Julia that has `Float32` performance that increases as the simulation progresses, while the `Float64` performance is constant in time. My suspicion is that is has to do with underflow in case of the `Float32` numbers. I have tested this by setting `set_zero_subnormals` at either `true` or `false`, but I do not see any difference in performance. What else could be the case here?
My code is at [https://github.com/Chiil/MicroHH.jl](https://github.com/Chiil/MicroHH.jl) and the test case can be run from the `cases/drycbl` directory using `julia --project -O3 -t1 drycbl_run.jl `. I could not reproduce this in a simple test program.
---
### Author:  [@gbaraldi](https://discourse.julialang.org/u/gbaraldi)
#### Post date: [October 22, 2021, 1:20pm UTC](https://discourse.julialang.org/t/float32-performance-fluctuating-over-time/70206/2 "2021-10-22T13:20:10Z")
I’m not sure what is happening in your case, but [https://github.com/milankl/Sherlogs.jl](https://github.com/milankl/Sherlogs.jl) is a very nice package for detecting these kinds of errors