-1

I have the standard mac mini m2 pro (10-core cpu, 512gb ssd, 16GB ram). I also have a mac studio m1 max (not ultra!) 32gb ram, upped to 1TB ssd. both run macos 13.2.1. I then brew install sysbench and

sysbench fileio --file-total-size=15G --file-test-mode=rndrw --time=300 --max-requests=0 run
sysbench fileio --file-total-size=15G --file-test-mode=rndrw --time=300 --max-requests=0 cleanup
sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)

the mini does

Running the test with following options:
Number of threads: 1
Initializing random number generator from current time


Extra file open flags: (none)
128 files, 120MiB each
15GiB total file size
Block size 16KiB
Number of IO requests: 0
Read/Write ratio for combined random IO test: 1.50
Periodic FSYNC enabled, calling fsync() each 100 requests.
Calling fsync() at the end of test, Enabled.
Using synchronous I/O mode
Doing random r/w test
Initializing worker threads...

Threads started!


File operations:
    reads/s:                      8668.41
    writes/s:                     5778.94
    fsyncs/s:                     18492.97

Throughput:
    read, MiB/s:                  135.44
    written, MiB/s:               90.30

General statistics:
    total time:                          300.0003s
    total number of events:              9881989

Latency (ms):
         min:                                    0.00
         avg:                                    0.03
         max:                                   20.48
         95th percentile:                        0.00
         sum:                               299269.99

Threads fairness:
    events (avg/stddev):           9881989.0000/0.00
    execution time (avg/stddev):   299.2700/0.00

sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)

the studio does

Running the test with following options:
Number of threads: 1
Initializing random number generator from current time


Extra file open flags: (none)
128 files, 120MiB each
15GiB total file size
Block size 16KiB
Number of IO requests: 0
Read/Write ratio for combined random IO test: 1.50
Periodic FSYNC enabled, calling fsync() each 100 requests.
Calling fsync() at the end of test, Enabled.
Using synchronous I/O mode
Doing random r/w test
Initializing worker threads...

Threads started!


File operations:
    reads/s:                      62863.67
    writes/s:                     41909.11
    fsyncs/s:                     134109.31

Throughput:
    read, MiB/s:                  982.24
    written, MiB/s:               654.83

General statistics:
    total time:                          300.0010s
    total number of events:              71664796

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                    8.05
         95th percentile:                        0.00
         sum:                               294415.57

Threads fairness:
    events (avg/stddev):           71664796.0000/0.00
    execution time (avg/stddev):   294.4156/0.00

sysbench 1.0.20 (using system LuaJIT 2.1.0-beta3)

Do I have something misconfigured on the m2 mini here? The studio pro has read throughput about 6-8 times faster than the mini. I have heard that the base mini m2 was slower, but this slow??

Improve this question
3
  • 1
    How do they compare doing 'real world' tasks? The 'scandal' suggests that the 256Gb models run at c. 1500 Mbps sequentially -- though they are supposedly better at random. I'd suggest you've just got a duff test.
    – benwiggy
    Mar 17 at 18:57
  • the reason why I ran this naive benchmark is that I became suspicious of some slower performance on my real-world app. (more sophisticated bacnhmarks try to guard against manipulation or shortcuts, but with this rather simple benchmark and the same os and manufacturer, this is less of a problem.) yes, I should add a sequential test, too, as it measures something quite different.
    – ivo Welch
    Mar 18 at 17:18
  • What about something like Blackmagic Disk Speed Tester?
    – benwiggy
    Mar 18 at 18:01

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .