Hey folks, so inspired by another thread I thought I'd do some benchmarking to see if choice of file system made a difference to file I/O performance. Software being used to benchmark is called AmorphousDiskMark - a very nice free clone of CrystalMark but for macOS.
Methodology:
The following was tested (APFS, APFS encrypted, HFS+, Fat, exFat). High and low utilization (disk space usage) were thrown into the mix to to see if that was a factor too. Initially two test runs on my system drive and all others on a Intel 660p 2TB in a TB3 enclosure. Results have 2 runs each to give you an idea of variability. macOS 10.14.6 (18G95) and hardware is a Macbook Pro 2018 13".
The results should really shock you. I don't understand what is going on so here's a summary:
1) Utilization did not seem to affect speed much.
2) 4k performance seems generally bad compared for an SSD with some online results. I guess this is to be expected?
3) Here's the zinger... ExFAT saw over 10 fold speed advantage over other file systems in 4k Write QD1. ExFat came in it at a blistering 360MB/s or 87890 IOPS whereas the group average was around 30MB/s.
To put this in perspective APFS saw 25MB/s, HFS+, 20MB/s and Fat32 60MB/s on 4k write, QD1.
I really can't explain whats going on. This seems to be truly remarkable if true, but it equally possible this is common knowledge. Another explanation is there's some sort of software bug
Well there you have it.
Full results posted below:
• Internal Apple SSD 2TB - 8% Free
• APFS Encrypted, 92% utilization
• Run 1
• Run 2
• Intel 660p over Thunderbolt 3 enclosure
• APFS Non-encyrpted, 98%utilization
• Run 1
• Run2
• APFS - 0% space utilization
• Run 1
• HFS+
• Run 1
• Run 2
• FAT32
• Run 1
• Run 2
• ExFAT - GUID
• Run 1
•
• Run 2
•
Methodology:
The following was tested (APFS, APFS encrypted, HFS+, Fat, exFat). High and low utilization (disk space usage) were thrown into the mix to to see if that was a factor too. Initially two test runs on my system drive and all others on a Intel 660p 2TB in a TB3 enclosure. Results have 2 runs each to give you an idea of variability. macOS 10.14.6 (18G95) and hardware is a Macbook Pro 2018 13".
The results should really shock you. I don't understand what is going on so here's a summary:
1) Utilization did not seem to affect speed much.
2) 4k performance seems generally bad compared for an SSD with some online results. I guess this is to be expected?
3) Here's the zinger... ExFAT saw over 10 fold speed advantage over other file systems in 4k Write QD1. ExFat came in it at a blistering 360MB/s or 87890 IOPS whereas the group average was around 30MB/s.
To put this in perspective APFS saw 25MB/s, HFS+, 20MB/s and Fat32 60MB/s on 4k write, QD1.
I really can't explain whats going on. This seems to be truly remarkable if true, but it equally possible this is common knowledge. Another explanation is there's some sort of software bug
Well there you have it.
Full results posted below:
• Internal Apple SSD 2TB - 8% Free
• APFS Encrypted, 92% utilization
• Run 1
• Run 2
• Intel 660p over Thunderbolt 3 enclosure
• APFS Non-encyrpted, 98%utilization
• Run 1
• Run2
• APFS - 0% space utilization
• Run 1
• HFS+
• Run 1
• Run 2
• FAT32
• Run 1
• Run 2
• ExFAT - GUID
• Run 1
•
• Run 2
•
Last edited: