Difference between revisions of "Talk:RPi Performance"
(Created page with "== Section on boot times? == How about a section with a list of boot times, for various distros, SD cards, with/without booting to GUI, etc? ~~~~") |
(→SD card benchmarking: new section) |
||
(One intermediate revision by one other user not shown) | |||
Line 1: | Line 1: | ||
== Section on boot times? == | == Section on boot times? == | ||
How about a section with a list of boot times, for various distros, SD cards, with/without booting to GUI, etc? [[User:FlyingLotus1983|FlyingLotus1983]] 02:56, 4 August 2012 (UTC) | How about a section with a list of boot times, for various distros, SD cards, with/without booting to GUI, etc? [[User:FlyingLotus1983|FlyingLotus1983]] 02:56, 4 August 2012 (UTC) | ||
+ | : According to me that seems a good idea. [[User:Jukimatori|Jukimatori]] 12:47, 5 August 2012 (UTC) | ||
+ | |||
+ | == SD card benchmarking == | ||
+ | |||
+ | New test procudure for SD cards needed? | ||
+ | |||
+ | Write tests should include time which is needed to write '''all''' data into card (=sync) | ||
+ | And before read tests file cache should be cleared. | ||
+ | |||
+ | Now people have two memory sizes which means that more ram is available for cache therefore results will be different with old and new model. | ||
+ | |||
+ | <pre> | ||
+ | dd if=/dev/zero of=~/test.tmp bs=500K count=1024 | ||
+ | time sync | ||
+ | sudo sh -c "echo 1 > /proc/sys/vm/drop_caches" | ||
+ | dd if=~/test.tmp of=/dev/null bs=500K count=1024 | ||
+ | sudo rm /mnt/test.tmp | ||
+ | time sync | ||
+ | |||
+ | 1024+0 records in | ||
+ | 1024+0 records out | ||
+ | 524288000 bytes (524 MB) copied, 44.8961 s, 11.7 MB/s | ||
+ | |||
+ | real 0m7.981s | ||
+ | user 0m0.000s | ||
+ | sys 0m0.000s | ||
+ | 1024+0 records in | ||
+ | 1024+0 records out | ||
+ | 524288000 bytes (524 MB) copied, 26.4597 s, 19.8 MB/s | ||
+ | |||
+ | real 0m1.555s | ||
+ | user 0m0.000s | ||
+ | sys 0m0.000s | ||
+ | ------------------- | ||
+ | 1024+0 records in | ||
+ | 1024+0 records out | ||
+ | 524288000 bytes (524 MB) copied, 43.6373 s, 12.0 MB/s | ||
+ | |||
+ | real 0m12.007s | ||
+ | user 0m0.000s | ||
+ | sys 0m0.000s | ||
+ | 1024+0 records in | ||
+ | 1024+0 records out | ||
+ | 524288000 bytes (524 MB) copied, 26.012 s, 20.2 MB/s | ||
+ | |||
+ | real 0m1.977s | ||
+ | user 0m0.000s | ||
+ | sys 0m0.000s | ||
+ | ------------------- | ||
+ | 1024+0 records in | ||
+ | 1024+0 records out | ||
+ | 524288000 bytes (524 MB) copied, 48.3281 s, 10.8 MB/s | ||
+ | |||
+ | real 0m8.183s | ||
+ | user 0m0.000s | ||
+ | sys 0m0.000s | ||
+ | 1024+0 records in | ||
+ | 1024+0 records out | ||
+ | 524288000 bytes (524 MB) copied, 26.8685 s, 19.5 MB/s | ||
+ | |||
+ | real 0m2.547s | ||
+ | user 0m0.000s | ||
+ | sys 0m0.000s | ||
+ | |||
+ | </pre> |
Latest revision as of 06:15, 2 November 2012
Section on boot times?
How about a section with a list of boot times, for various distros, SD cards, with/without booting to GUI, etc? FlyingLotus1983 02:56, 4 August 2012 (UTC)
- According to me that seems a good idea. Jukimatori 12:47, 5 August 2012 (UTC)
SD card benchmarking
New test procudure for SD cards needed?
Write tests should include time which is needed to write all data into card (=sync) And before read tests file cache should be cleared.
Now people have two memory sizes which means that more ram is available for cache therefore results will be different with old and new model.
dd if=/dev/zero of=~/test.tmp bs=500K count=1024 time sync sudo sh -c "echo 1 > /proc/sys/vm/drop_caches" dd if=~/test.tmp of=/dev/null bs=500K count=1024 sudo rm /mnt/test.tmp time sync 1024+0 records in 1024+0 records out 524288000 bytes (524 MB) copied, 44.8961 s, 11.7 MB/s real 0m7.981s user 0m0.000s sys 0m0.000s 1024+0 records in 1024+0 records out 524288000 bytes (524 MB) copied, 26.4597 s, 19.8 MB/s real 0m1.555s user 0m0.000s sys 0m0.000s ------------------- 1024+0 records in 1024+0 records out 524288000 bytes (524 MB) copied, 43.6373 s, 12.0 MB/s real 0m12.007s user 0m0.000s sys 0m0.000s 1024+0 records in 1024+0 records out 524288000 bytes (524 MB) copied, 26.012 s, 20.2 MB/s real 0m1.977s user 0m0.000s sys 0m0.000s ------------------- 1024+0 records in 1024+0 records out 524288000 bytes (524 MB) copied, 48.3281 s, 10.8 MB/s real 0m8.183s user 0m0.000s sys 0m0.000s 1024+0 records in 1024+0 records out 524288000 bytes (524 MB) copied, 26.8685 s, 19.5 MB/s real 0m2.547s user 0m0.000s sys 0m0.000s