# Image IO and processing efficiency: sequential versus bulk

**URL:** <https://discourse.julialang.org/t/image-io-and-processing-efficiency-sequential-versus-bulk/48975>\
**Category:** Performance\
**Tags:** images, performance, memory-allocation, optimization, time\
**Created:** [October 25, 2020, 12:51pm UTC](https://discourse.julialang.org/t/image-io-and-processing-efficiency-sequential-versus-bulk/48975 "2020-10-25T12:51:05Z")\
**Posts on this page:** 1\
**Showing post:** 8

<div class="post-metadata">

**Author:** ![libensvivit](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/libensvivit/32/19006_2.png) [@libensvivit](https://discourse.julialang.org/u/libensvivit)\
**Post date:** [October 26, 2020, 12:54am UTC](https://discourse.julialang.org/t/image-io-and-processing-efficiency-sequential-versus-bulk/48975/8 "2020-10-26T00:54:12Z")

</div>

I’m also trying to read image files as fast as possible, you can check my question [here](https://discourse.julialang.org/t/plotting-images-on-gpu-with-arrayfire/48676)

Loading 10k images with Images.load() on my computer takes:  
`21.982461 seconds (2.49 M allocations: 991.897 MiB, 0.47% gc time)`

Loading the same images with ArrayFire.load\_image() function gives:  
`7.332682 seconds (229.77 k allocations: 14.024 MiB)`

ArrayFire stores the images on GPU so I’m not confident with memory allocation but it is surely time-efficient. I am not using bulk reading in my algorithm but load and process is fast enough combined with @distributed parallelism.

---

_[View the full topic](https://discourse.julialang.org/t/image-io-and-processing-efficiency-sequential-versus-bulk/48975)._